[{"case_id":"AIBD-0214","case_url":"records/AIBD-0214.html","measure":"Algebra proficiency.","reported_result":"scored 0.21 standard deviations higher in the second year","baseline":"Comparison schools in the same implementation year","result":"Second high-school cohort: +0.21 standard deviations; p = .04 before correction, not significant at the Clearinghouse’s corrected threshold of .025","baseline_value":null,"result_value":null,"unit":null,"effect":"The revised high-school estimate was −0.07 standard deviations in year one and +0.21 in year two.","design":"Researchers assigned matched pairs of schools to Cognitive Tutor or their usual curriculum.","sample":"The trial included 73 high schools and 74 middle schools. The revised analysis excluded students without observed baseline tests.","uncertainty":"The second-year estimate depended on adjustment for student characteristics. Its p-value of .04 did not meet the Clearinghouse’s corrected threshold of .025.","comparison_basis":"measured","source_urls":["https://ies.ed.gov/ncee/wwc/Docs/InterventionReports/wwc_cognitivetutor_122116.pdf","https://journals.sagepub.com/doi/10.3102/0162373713507480","https://www.rand.org/content/dam/rand/pubs/working_papers/WR1000/WR1050/RAND_WR1050.pdf","https://link.springer.com/content/pdf/10.3758/BF03194060.pdf","https://www.rand.org/content/dam/rand/pubs/technical_reports/2012/RAND_TR1171-1.pdf","https://educationaldatamining.org/EDM2024/proceedings/2024.EDM-short-papers.52/index.html","https://arxiv.org/abs/1808.02528","https://arxiv.org/abs/1802.08616"],"source_refs":["AIBD-0214-source-0001","AIBD-0214-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The software’s separate contribution was not estimated.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Schools continued the algebra curriculum that they normally used.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Researchers assigned matched pairs of schools to Cognitive Tutor or their usual curriculum.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"The revised high-school estimate was −0.07 standard deviations in year one and +0.21 in year two.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Students generally used the tutor for about 40% of class time during the school year.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"AIBD-0214-E01","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Students took standardized algebra readiness and proficiency tests. Researchers adjusted the end-of-year comparison for baseline scores and other characteristics.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Algebra proficiency.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The 2024 reanalysis provides public code. Its full replication data are withheld to protect the participating schools’ identities.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"The trial included 73 high schools and 74 middle schools. The revised analysis excluded students without observed baseline tests.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The second-year estimate depended on adjustment for student characteristics. Its p-value of .04 did not meet the Clearinghouse’s corrected threshold of .025.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0214-reported-measure","evaluation_data":null},{"case_id":"AIBD-0168","case_url":"records/AIBD-0168.html","measure":"cut mean wind forecast error from about 18% to 10.8% by 2014; Xcel estimated US$49m saved","reported_result":"cut mean wind forecast error from about 18% to 10.8% by 2014; Xcel estimated US$49m saved","baseline":"about 18 percent for 18 to 42 hours ahead","result":"10.8 percent","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://spectrum.ieee.org/taming-wind-power-with-better-forecasts","https://www.eenews.net/articles/xcel-learns-to-forecast-wind-then-shuts-coal-plants/","https://news.ucar.edu/6460/wind-power-wisdom","https://www.osti.gov/servlets/purl/1027161","https://www.technologyreview.com/technology/smart-wind-and-solar-power/"],"source_refs":["AIBD-0168-source-0001","AIBD-0168-source-0002","AIBD-0168-source-0003","AIBD-0168-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Xcel's wind forecasts in 2008, before NCAR's system","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: about 18 percent for 18 to 42 hours ahead. Result: 10.8 percent.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"cut mean wind forecast error from about 18% to 10.8% by 2014; Xcel estimated US$49m saved","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0168-reported-measure","evaluation_data":null},{"case_id":"AIBD-0054","case_url":"records/AIBD-0054.html","measure":"percentage change in phantom limb pain.","reported_result":"after ten weeks near-absent pain developing into completely pain-free periods","baseline":"constant burning pain of an intensity of 3 on a scale from 0 to 10 (SF-MGPQ)","result":"gradually reduced to complete pain-free periods","baseline_value":null,"result_value":null,"unit":null,"effect":"pain fell 64.5% with phantom motor execution and 68.2% with imagery; 71% and 68% of each group reached a clinically meaningful reduction.","design":"randomized.","sample":"81 people with phantom limb pain randomized at 9 outpatient clinics in Sweden, the Netherlands, Slovenia, Ireland, Germany, Canada and the United States.","uncertainty":"No untreated control group, so the shared effect of training in extended reality cannot be separated from natural change. The method's developer is senior author and two authors list Integrum AB, which part-funded the original case report.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC3935120/","https://europepmc.org/abstract/MED/27916234","https://pmc.ncbi.nlm.nih.gov/articles/PMC11808706/","https://pmc.ncbi.nlm.nih.gov/articles/PMC9776775/"],"source_refs":["AIBD-0054-source-0004"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"none detected; the authors found no evidence that decoded execution beat imagery.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"phantom motor imagery guided by extended reality, with no decoding of muscle signals and the same equipment.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"pain fell 64.5% with phantom motor execution and 68.2% with imagery; 71% and 68% of each group reached a clinically meaningful reduction.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Change in the Pain Rating Index from baseline to the end of treatment, and the rate of clinically meaningful pain reduction of more than 50%.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"percentage change in phantom limb pain.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"81 people with phantom limb pain randomized at 9 outpatient clinics in Sweden, the Netherlands, Slovenia, Ireland, Germany, Canada and the United States.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"No untreated control group, so the shared effect of training in extended reality cannot be separated from natural change. The method's developer is senior author and two authors list Integrum AB, which part-funded the original case report.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0054-reported-measure","evaluation_data":null},{"case_id":"AIBD-0008","case_url":"records/AIBD-0008.html","measure":"Rear-end striking crash involvement rates, with and without injuries.","reported_result":"reduced rear-end striking crash involvement rates by 27%, 43% and 50%","baseline":"the same models where the optional systems were not purchased","result":"reduced rear-end striking crash involvement rates by 27%, 43%, and 50%, respectively","baseline_value":null,"result_value":null,"unit":null,"effect":"Rear-end striking crash rates fell 27% (FCW alone), 43% (low-speed AEB) and 50% (FCW with AEB); injury crash rates fell 20%, 45% and 56%.","design":"controlled observational.","sample":"Passenger vehicles in 22 U.S. states, 2010 to 2014; models with optional FCW or FCW with AEB, and Volvo 2011-2012 S60 and 2010-2012 XC60 with standard low-speed AEB.","uncertainty":"Owners chose optional systems, so buyers may differ from other drivers; the low-speed AEB comparison uses different models; FCW with AEB raised rates of being struck from behind by 20%.","comparison_basis":"measured","source_urls":["https://www.iihs.org/research-areas/bibliography/ref/2111","https://www.reginfo.gov/public/do/eoDownloadDocument?pubId=&eodoc=true&documentID=219043","https://www.federalregister.gov/documents/2024/05/09/2024-09054/federal-motor-vehicle-safety-standards-automatic-emergency-braking-systems-for-light-vehicles","https://www.federalregister.gov/documents/2023/06/13/2023-11863/federal-motor-vehicle-safety-standards-automatic-emergency-braking-systems-for-light-vehicles","https://natlawreview.com/article/road-ahead-fmvss-127-whither-automatic-emergency-braking-mandate"],"source_refs":["https://www.iihs.org/research-areas/bibliography/ref/2111"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The same models without the optional system; for low-speed AEB, other luxury midsize cars and SUVs without it.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Rear-end striking crash rates fell 27% (FCW alone), 43% (low-speed AEB) and 50% (FCW with AEB); injury crash rates fell 20%, 45% and 56%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Insured vehicle years in the study states.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Poisson regression of police-reported crash involvements per insured vehicle year, controlling for other factors that affect crash risk.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Rear-end striking crash involvement rates, with and without injuries.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Passenger vehicles in 22 U.S. states, 2010 to 2014; models with optional FCW or FCW with AEB, and Volvo 2011-2012 S60 and 2010-2012 XC60 with standard low-speed AEB.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Owners chose optional systems, so buyers may differ from other drivers; the low-speed AEB comparison uses different models; FCW with AEB raised rates of being struck from behind by 20%.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0008-reported-measure","evaluation_data":null},{"case_id":"AIBD-0670","case_url":"records/AIBD-0670.html","measure":"Export quantity and revenue.","reported_result":"raised exports on eBay by 10.9% in the published study","baseline":"Exports with Bing Translator before May 2014","result":"10.9% more exports (published); 17.5% to 20.9% more (working paper)","baseline_value":null,"result_value":null,"unit":null,"effect":"10.9% more exports (published); 17.5% to 20.9% in quantity and 13.1% to 17% in revenue (working paper).","design":"quasi-experimental.","sample":"U.S. exports on eBay to Latin America and other countries, six and twelve months around the change.","uncertainty":"One platform; nonrandom rollout timing; eBay supplied the data; one author interned at eBay in 2015; the published estimate is smaller than the working paper estimate.","comparison_basis":"measured","source_urls":["https://doi.org/10.1287/mnsc.2019.3388","https://www.nber.org/system/files/working_papers/w24917/w24917.pdf"],"source_refs":["AIBD-0670-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"About a tenth to a fifth more exports over the older machine translation.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Other export destinations on eBay and offline exports.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"10.9% more exports (published); 17.5% to 20.9% in quantity and 13.1% to 17% in revenue (working paper).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Machine-translated listing titles.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Export quantity and revenue from eBay transaction data.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Export quantity and revenue.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Data not public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"U.S. exports on eBay to Latin America and other countries, six and twelve months around the change.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One platform; nonrandom rollout timing; eBay supplied the data; one author interned at eBay in 2015; the published estimate is smaller than the working paper estimate.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0670-reported-measure","evaluation_data":null},{"case_id":"AIBD-0190","case_url":"records/AIBD-0190.html","measure":"Study 1: Share of establishments with critical violations found in the first half of the work; average days earlier. Study 2: Days earlier to find establishments with a critical violation; monthly hit rates by violation code before and after January 2015.","reported_result":"would have found 178 (69%) in the first half of the work, on average 7.5 days earlier","baseline":"55 percent (141 establishments) of those with critical violations found in the first month during normal operations","result":"69 percent (178 establishments) found in the first half of work; violations found on average 7 and a half days earlier","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: 69 percent found in the first half against 55 percent; violations found on average 7.5 days earlier. Study 2: 7.438 days earlier on average in simulation, standard deviation 25.156 days; hit rates rose after 2015 for several violation codes.","design":"Study 1: simulation. Study 2: simulation.","sample":"Study 1: 1,637 canvass inspections by the Chicago Department of Public Health, September and October 2014; 258 had at least one critical violation. Study 2: The city's 1,637-inspection test set; 50,462 canvass and 17,088 complaint inspections from the public data.","uncertainty":"Study 1: No inspection was changed; the simulation assumes that results would be the same on an earlier date, and the city reports no confidence interval. Written by the city that built and uses the model. Study 2: The authors could not observe when establishments would have been inspected without the model, and the city does not release inspector assignments or inspection cycles. Presented at a symposium; not a journal article.","comparison_basis":"measured","source_urls":["https://chicago.github.io/food-inspections-evaluation/","http://datasmart.ash.harvard.edu/news/article/delivering-faster-results-with-food-inspection-forecasting-631","https://arxiv.org/abs/1910.04906"],"source_refs":["AIBD-0190-source-0001","https://arxiv.org/html/1910.04906"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: The whole difference is attributed to the model's ordering, because the same inspections are re-ordered. Study 2: The authors argue that the post-2015 rise cannot be attributed to the model, because restaurant turnover and stricter inspection could explain it.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: The department's usual inspection order in the same two months. Study 2: The usual inspection order in the test set; hit rates before January 2015.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: simulation. Study 2: simulation.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 69 percent found in the first half against 55 percent; violations found on average 7.5 days earlier. Study 2: 7.438 days earlier on average in simulation, standard deviation 25.156 days; hit rates rose after 2015 for several violation codes.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Simulation only. Study 2: Model in live use from January 2015.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: simulation. Study 2: simulation.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Double-blind retrodiction: inspections done in the usual order were re-ordered by the model's predicted risk, and the timing of critical-violation findings was compared. Study 2: Independent re-run of the city's released model and test set, plus a before and after comparison of monthly violation hit rates in public inspection records from January 2010 to June 2018.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Share of establishments with critical violations found in the first half of the work; average days earlier. Study 2: Days earlier to find establishments with a critical violation; monthly hit rates by violation code before and after January 2015.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Code and data released by the City of Chicago on GitHub. Study 2: Uses the city's public code and data on GitHub and the Chicago data portal.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 1,637 canvass inspections by the Chicago Department of Public Health, September and October 2014; 258 had at least one critical violation. Study 2: The city's 1,637-inspection test set; 50,462 canvass and 17,088 complaint inspections from the public data.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: No inspection was changed; the simulation assumes that results would be the same on an earlier date, and the city reports no confidence interval. Written by the city that built and uses the model. Study 2: The authors could not observe when establishments would have been inspected without the model, and the city does not release inspector assignments or inspection cycles. Presented at a symposium; not a journal article.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0190-reported-measure","evaluation_data":null},{"case_id":"AIBD-0174","case_url":"records/AIBD-0174.html","measure":"linked a third company to a suspected money laundering network","reported_result":"linked a third company to a suspected money laundering network","baseline":"manual approaches to reviewing documentation for signatories, which the report calls the only means to make these connections","result":"the tool linked Company C to Company A through shared phone numbers and email addresses","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.hkma.gov.hk/media/eng/doc/key-information/guidelines-and-circular/2021/20210121e2a1.pdf","https://www.hkma.gov.hk/eng/news-and-media/press-releases/2021/01/20210121-3/"],"source_refs":["AIBD-0174-source-0001","AIBD-0174-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"even if 10% were unrecoverable the tool contributed around USD 50 million in savings","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"even if 10% were unrecoverable the tool contributed around USD 50 million in savings","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0174-reported-measure","evaluation_data":null},{"case_id":"AIBD-0176","case_url":"records/AIBD-0176.html","measure":"about 8 percent of A-rated and 28 percent of B-rated borrowers had FICO scores below 680","reported_result":"about 8 percent of A-rated and 28 percent of B-rated borrowers had FICO scores below 680","baseline":"classified as subprime by traditional criteria","result":"slotted into “better” loan grades, which allowed them to get lower-priced credit","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.philadelphiafed.org/-/media/frbp/assets/working-papers/2018/wp18-15r.pdf","https://web.archive.org/web/20250613021044/https://onlinelibrary.wiley.com/doi/10.1111/fima.12295","https://www.philadelphiafed.org/-/media/frbp/assets/working-papers/2023/wp23-26.pdf"],"source_refs":["AIBD-0176-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"classification of borrowers by traditional credit criteria","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: classified as subprime by traditional criteria. Result: slotted into “better” loan grades, which allowed them to get lower-priced credit.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"about 8 percent of A-rated and 28 percent of B-rated borrowers had FICO scores below 680","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0176-reported-measure","evaluation_data":null},{"case_id":"AIBD-0186","case_url":"records/AIBD-0186.html","measure":"$654.8 million in inappropriate Medicare payments identified or prevented in 2015","reported_result":"$654.8 million in inappropriate Medicare payments identified or prevented in 2015","baseline":"identified savings from the previous year ($454.0 million in Table 1)","result":"$654.8 million","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/20160824134124id_/https://www.cms.gov/about-cms/components/cpi/downloads/fraud-prevention-system-return-on-investment-fourth-implementation-year-2015.pdf","https://web.archive.org/web/20160824134124/https://www.cms.gov/about-cms/components/cpi/downloads/fraud-prevention-system-return-on-investment-fourth-implementation-year-2015.pdf","https://web.archive.org/web/2026/https://www.gao.gov/products/gao-17-710"],"source_refs":["AIBD-0186-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the FPS's identified savings in the previous year, 2014","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: identified savings from the previous year ($454.0 million in Table 1). Result: $654.8 million.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"$654.8 million in inappropriate Medicare payments identified or prevented in 2015","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0186-reported-measure","evaluation_data":null},{"case_id":"AIBD-0196","case_url":"records/AIBD-0196.html","measure":"a US$1 million fine plus a US$1 million goodwill grant followed","reported_result":"a US$1 million fine plus a US$1 million goodwill grant followed","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://grist.org/science/this-new-digital-tool-could-help-fight-overfishing-around-the-world/","https://www.washingtonpost.com/news/energy-environment/wp/2016/09/15/from-space-a-new-effort-to-crack-down-on-illegal-fishing-across-the-globe/","https://www.bbc.com/future/article/20180607-how-to-spot-the-secretive-activities-of-rogue-fishing-boats","https://globalfishingwatch.org/transparency/indonesia-shares-vms-with-global-fishing-watch/","https://web.archive.org/web/2026/https://pmc.ncbi.nlm.nih.gov/articles/PMC6431217/"],"source_refs":["AIBD-0196-source-0002","AIBD-0196-source-0001","AIBD-0196-source-0003","AIBD-0196-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a US$1 million fine plus a US$1 million goodwill grant followed","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a US$1 million fine plus a US$1 million goodwill grant followed","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0196-reported-measure","evaluation_data":null},{"case_id":"AIBD-0635","case_url":"records/AIBD-0635.html","measure":"Dollar value of confirmed identity-theft returns selected only by each system.","reported_result":"selected $1.88 billion, or 47.8%, of $3.92 billion in confirmed identity-theft refunds","baseline":"$60 million (1.5%) of confirmed identity-theft refunds selected only by the older system","result":"$1.88 billion (47.8%) selected only by the Return Review Program models","baseline_value":null,"result_value":null,"unit":null,"effect":"$1.88 billion (47.8%) unique to the Return Review Program models against $60 million (1.5%) unique to the older system.","design":"other.","sample":"700,560 confirmed identity-theft returns, 2015 filing season.","uncertainty":"Figures come from IRS data; selections by manual processes are not counted; parts of the report are redacted; no comparison with no screening.","comparison_basis":"measured","source_urls":["https://www.tigta.gov/sites/default/files/reports/2022-02/201720080fr.pdf","https://www.gao.gov/assets/gao-18-544.pdf"],"source_refs":["AIBD-0635-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated from the program's business rules.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Electronic Fraud Detection System and the Dependent Database.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"$1.88 billion (47.8%) unique to the Return Review Program models against $60 million (1.5%) unique to the older system.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"All electronically filed individual returns claiming refunds.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Audit of IRS filing-season data on which detection systems selected each confirmed identity-theft return.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Dollar value of confirmed identity-theft returns selected only by each system.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"None public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"700,560 confirmed identity-theft returns, 2015 filing season.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Figures come from IRS data; selections by manual processes are not counted; parts of the report are redacted; no comparison with no screening.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0635-reported-measure","evaluation_data":null},{"case_id":"AIBD-0188","case_url":"records/AIBD-0188.html","measure":"Demerits per inspection; venues graded C or worse.","reported_result":"found 11 venues graded C or worse versus 7 for controls","baseline":"6 demerits per inspection and 7 venues graded C or worse in the control arm","result":"9 demerits per inspection and 11 venues graded C or worse in the nEmesis arm","baseline_value":null,"result_value":null,"unit":null,"effect":"9 against 6 demerits per inspection (p = 0.019); 11 against 7 venues graded C or worse.","design":"quasi-experimental.","sample":"142 food venues across the Las Vegas metropolitan area, half chosen by nEmesis and half by the standard protocol, January to March 2015.","uncertainty":"Small sample over three months; SNHD staff and the developers wrote the paper together; the prevented illnesses and hospitalizations are model estimates from CDC outbreak averages.","comparison_basis":"measured","source_urls":["https://ojs.aaai.org/aimagazine/index.php/aimagazine/article/view/2711","https://ojs.aaai.org/index.php/AAAI/article/view/19072"],"source_refs":["AIBD-0188-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Venues chosen under the standard inspection protocol.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"9 against 6 demerits per inspection (p = 0.019); 11 against 7 venues graded C or worse.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Demerits and letter grades recorded by SNHD environmental health specialists in standard inspections.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Demerits per inspection; venues graded C or worse.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"142 food venues across the Las Vegas metropolitan area, half chosen by nEmesis and half by the standard protocol, January to March 2015.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small sample over three months; SNHD staff and the developers wrote the paper together; the prevented illnesses and hospitalizations are model estimates from CDC outbreak averages.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0188-reported-measure","evaluation_data":null},{"case_id":"AIBD-0057","case_url":"records/AIBD-0057.html","measure":"Study 1: Number of the 10 tasks completed. Study 2: Questionnaire subscale scores, above all near activities and reading.","reported_result":"improved on a 10-item daily-living test from 2.5 to 9.5","baseline":"a mean of 2.5 of 10 daily-living tasks completed","result":"a mean of 9.5 of 10 tasks with the device at the first visit","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: Mean 2.5 tasks without the device, 9.5 with it at the first visit and 9.8 after a week of home use. Study 2: Significant gains on the near-activities and reading measures; 3 of 20 patients kept using the device afterward.","design":"Study 1: before/after. Study 2: before/after.","sample":"Study 1: 12 patients with best-corrected acuity of 20/200 or worse in a US ophthalmology department, 1 July to 30 September 2015. Study 2: 20 Dutch patients with retinitis pigmentosa or cone-rod dystrophy and acuity of 20/400 or worse.","uncertainty":"Study 1: Small sample, no control group, a test that is not validated and one week of follow-up. Study 2: Small sample, no control group and self-reported measures; 17 of 20 patients stopped using the device and most often cited other text-to-speech tools such as Seeing AI, the price or missing features.","comparison_basis":"measured","source_urls":["https://europepmc.org/abstract/MED/27148909","https://pmc.ncbi.nlm.nih.gov/articles/PMC10863346/","https://europepmc.org/abstract/MED/36645535","https://pmc.ncbi.nlm.nih.gov/articles/PMC13173725/","https://pmc.ncbi.nlm.nih.gov/articles/PMC9292690/"],"source_refs":["AIBD-0057-source-0002","AIBD-0057-source-0005"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: The full gain is attributed to the device, since the comparison is the same patients without it.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Same patients using best-corrected visual acuity alone. Study 2: Same patients at baseline before using the device.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: before/after. Study 2: before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Mean 2.5 tasks without the device, 9.5 with it at the first visit and 9.8 after a week of home use. Study 2: Significant gains on the near-activities and reading measures; 3 of 20 patients kept using the device afterward.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: One clinic session and one week of home use. Study 2: About five weeks of use.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: before/after. Study 2: before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Non-validated 10-item test that simulates daily-living tasks. Study 2: Dutch NEI-VFQ, Participation and Activity Inventory and an OrCam Function Questionnaire at baseline and after about five weeks.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Number of the 10 tasks completed. Study 2: Questionnaire subscale scores, above all near activities and reading.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 12 patients with best-corrected acuity of 20/200 or worse in a US ophthalmology department, 1 July to 30 September 2015. Study 2: 20 Dutch patients with retinitis pigmentosa or cone-rod dystrophy and acuity of 20/400 or worse.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Small sample, no control group, a test that is not validated and one week of follow-up. Study 2: Small sample, no control group and self-reported measures; 17 of 20 patients stopped using the device and most often cited other text-to-speech tools such as Seeing AI, the price or missing features.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0057-reported-measure","evaluation_data":null},{"case_id":"AIBD-0105","case_url":"records/AIBD-0105.html","measure":"decreased the probability of deforestation in Africa by 18 percent","reported_result":"decreased the probability of deforestation in Africa by 18 percent","baseline":"average 2011–2016 levels","result":"probability of deforestation in Africa decreased by 18%; no effect on other continents","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.nature.com/articles/s41558-020-00956-w","https://iopscience.iop.org/article/10.1088/1748-9326/11/3/034008","https://globalnaturewatch.org/blog/data-and-tools/integrated-deforestation-alerts/"],"source_refs":["AIBD-0105-source-0001","AIBD-0105-source-0002","AIBD-0105-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"average 2011–2016 deforestation levels, with areas lacking alert subscribers as the counterfactual across 22 tropical countries","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: average 2011–2016 levels. Result: probability of deforestation in Africa decreased by 18%; no effect on other continents.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"decreased the probability of deforestation in Africa by 18 percent","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0105-reported-measure","evaluation_data":null},{"case_id":"AIBD-0166","case_url":"records/AIBD-0166.html","measure":"Pipe-material prediction, inspection success, and modeled cost per replacement.","reported_result":"held-out prediction accuracy was 91.6%","baseline":"Estimated US$5,818 per replacement","result":"Estimated US$5,196 per replacement","baseline_value":null,"result_value":null,"unit":null,"effect":"The held-out prediction accuracy was 91.6%. The simulated strategy lowered estimated cost per replacement by 10.7%.","design":"Researchers tested predictions on held-out homes and compared inspection strategies in simulations.","sample":"The database covered 55,893 parcels. The main accuracy test held out 1,606 homes with known pipe materials.","uncertainty":"Known pipe materials came from a selected set of inspected homes. The simulated costs omitted travel and other logistical constraints.","comparison_basis":"assumed","source_urls":["https://arxiv.org/abs/1806.10692","https://www.theatlantic.com/technology/archive/2019/01/how-machine-learning-found-flints-lead-pipes/578692/","https://www.nrdc.org/sites/default/files/order-amending-settlement-agreement-flint-20190326.pdf","https://www.michiganpublic.org/news/2019-04-25/in-flint-algorithm-predicts-lead-pipe-locations-as-replacement-work-continues","https://pmc.ncbi.nlm.nih.gov/articles/PMC11160116/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11152035/","https://www.wired.com/story/algorithm-helping-community-detect-lead-pipes/","https://web.archive.org/web/20250702214854/https://michiganadvance.com/2025/07/02/flint-completes-lead-pipe-replacement-11-years-after-beginning-of-water-crisis/"],"source_refs":["AIBD-0166-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The estimated cost improvement came from a simulation. The field results do not isolate the model from crews, funding, and policy decisions.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The cost simulation compared the proposed strategy with the sequence of work in the city’s records.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Researchers tested predictions on held-out homes and compared inspection strategies in simulations.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"The held-out prediction accuracy was 91.6%. The simulated strategy lowered estimated cost per replacement by 10.7%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Predictions were updated as crews recorded new inspections and replacements.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"AIBD-0166-E01, AIBD-0166-E02","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Excavation results supplied the pipe-material labels. Validation and simulated work sequences used those records.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Pipe-material prediction, inspection success, and modeled cost per replacement.","inferred":false}},"evaluations[].reproduction_materials":{"state":"not_reported","review_state":"reviewed","display":{"answer":"The paper describes the algorithms and simulations. It does not supply all settings needed to reproduce the deployed system.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"The database covered 55,893 parcels. The main accuracy test held out 1,606 homes with known pipe materials.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Known pipe materials came from a selected set of inspected homes. The simulated costs omitted travel and other logistical constraints.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0166-reported-measure","evaluation_data":null},{"case_id":"AIBD-0212","case_url":"records/AIBD-0212.html","measure":"Math and Hindi test scores.","reported_result":"0.37 standard deviations in math and 0.23 in Hindi over 4.5 months","baseline":"Usual schooling and activities","result":"Math +0.37 standard deviations; Hindi +0.23 over 4.5 months","baseline_value":null,"result_value":null,"unit":null,"effect":"The voucher offer raised math scores by 0.37 standard deviations and Hindi scores by 0.23.","design":"A lottery assigned students to a free voucher or a comparison group. Researchers estimated the effect of the offer.","sample":"There were 314 voucher recipients and 305 controls. Endline participation was 85% and 90%, respectively.","uncertainty":"Participants volunteered for the study. Endline attrition differed between groups, and longer-term attainment was not measured.","comparison_basis":"measured","source_urls":["https://www.aeaweb.org/articles?id=10.1257/aer.20171112","https://cdep.sipa.columbia.edu/sites/cdep.sipa.columbia.edu/files/content/Mindspark_Rajasthan%20(7%20October%202025).pdf"],"source_refs":["AIBD-0212-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The design did not separate adaptive software from additional teaching and practice.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Students who were not offered a voucher continued their usual schooling and activities.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"A lottery assigned students to a free voucher or a comparison group. Researchers estimated the effect of the offer.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"The voucher offer raised math scores by 0.37 standard deviations and Hindi scores by 0.23.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Average attendance was 58%, or about 50 of the 86 available days.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"AIBD-0212-E01","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Independent paper tests assessed math and Hindi before and after the program. Scores were placed on comparable scales.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Math and Hindi test scores.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Replication data are deposited with the American Economic Association at openICPSR.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"There were 314 voucher recipients and 305 controls. Endline participation was 85% and 90%, respectively.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Participants volunteered for the study. Endline attrition differed between groups, and longer-term attainment was not measured.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0212-reported-measure","evaluation_data":null},{"case_id":"AIBD-0049","case_url":"records/AIBD-0049.html","measure":"Genetic diagnoses were found in 25% of probands, rising to 35% for monogenic disorders","reported_result":"Genetic diagnoses were found in 25% of probands, rising to 35% for monogenic disorders","baseline":"A total of 322 (66%) of the 490 diagnoses that were based on SNVs or indels from the genomes were made with the virtual panel–based pipeline","result":"diagnoses were detected in 77% of the top-ranked candidate variants, in 86% of the top three candidates, and in 88% of the top five candidates","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC7613219/","https://doi.org/10.1186/s13073-025-01546-1","https://pmc.ncbi.nlm.nih.gov/articles/PMC11814671/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11655964/"],"source_refs":["AIBD-0049-source-0001","AIBD-0049-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the virtual panel-based pipeline (applied gene panels) alone","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: A total of 322 (66%) of the 490 diagnoses that were based on SNVs or indels from the genomes were made with the virtual panel–based pipeline. Result: diagnoses were detected in 77% of the top-ranked candidate variants, in 86% of the top three candidates, and in 88% of the top five candidates.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Genetic diagnoses were found in 25% of probands, rising to 35% for monogenic disorders","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0049-reported-measure","evaluation_data":null},{"case_id":"AIBD-0265","case_url":"records/AIBD-0265.html","measure":"Selection accuracy, spelling speed, time with the system available at home, satisfaction, and persistence of usable control.","reported_result":"seven years of independent at-home use","baseline":"Eye tracker PIADS scores: competence 1.3, adaptability 1.5, self-esteem 0.6","result":"Implant PIADS scores: competence 1.1, adaptability 2.2, self-esteem 1.0","baseline_value":null,"result_value":null,"unit":null,"effect":"Initial spelling accuracy averaged 89%. With word prediction, she selected about two letters per minute. Home use continued for nearly seven years.","design":"A prospective single-participant study, followed by a longitudinal report on the same woman.","sample":"One participant at UMC Utrecht and in her home. Repeated task runs are observations from that same person.","uncertainty":"The result concerns one person. Home logs show availability, not continuous active communication. Declining control coincided with ALS progression, without proving a single cause.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC5326682/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11395392/","https://doi.org/10.33540/2877","https://www.scientificamerican.com/article/wireless-brain-implant-allows-ldquo-locked-in-rdquo-woman-to-communicate/"],"source_refs":["AIBD-0265-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The study demonstrates the whole implanted interface. It does not isolate a machine-learning contribution, and its decoder is described through signal processing and fitted thresholds.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"There was no control group. The participant also used an eye tracker until eye movement became unreliable.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"A prospective single-participant study, followed by a longitudinal report on the same woman.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Initial spelling accuracy averaged 89%. With word prediction, she selected about two letters per minute. Home use continued for nearly seven years.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"The initial report describes 67 two-hour research sessions over 262 days. Home use without investigators began on day 197 after implantation.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"NCT02224469","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Researchers tested clicks and spelling, collected satisfaction ratings, logged when the home system was available, and monitored brain signals and imaging.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Selection accuracy, spelling speed, time with the system available at home, satisfaction, and persistence of usable control.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The original paper links two example task runs in Dryad. The follow-up supplement describes signal analysis. The thesis documents click thresholds and reproduces the follow-up supplement.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One participant at UMC Utrecht and in her home. Repeated task runs are observations from that same person.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The result concerns one person. Home logs show availability, not continuous active communication. Declining control coincided with ALS progression, without proving a single cause.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0265-reported-measure","evaluation_data":null},{"case_id":"AIBD-0058","case_url":"records/AIBD-0058.html","measure":"gained 4.58 points more on the Vineland socialization subscale than controls (P = .005)","reported_result":"gained 4.58 points more on the Vineland socialization subscale than controls (P = .005)","baseline":null,"result":"mean [SD] treatment impact, 4.58 [1.62]; P = .005","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://europepmc.org/abstract/MED/30907929","https://europepmc.org/article/PMC/PMC6550272","https://pmc.ncbi.nlm.nih.gov/articles/PMC11659516/","https://www.thetransmitter.org/spectrum/tech-firms-superpower-glass-autism-not-super-experts-say/"],"source_refs":["AIBD-0058-source-0002","AIBD-0058-source-0001","AIBD-0058-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"applied behavioral analysis therapy alone (treatment as usual)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Result: mean [SD] treatment impact, 4.58 [1.62]; P = .005.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"gained 4.58 points more on the Vineland socialization subscale than controls (P = .005)","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0058-reported-measure","evaluation_data":null},{"case_id":"AIBD-0215","case_url":"records/AIBD-0215.html","measure":"Enrollment, financial-aid tasks, and other requirements for entering college.","reported_result":"raised enrollment by 3.3 percentage points among students already committed to Georgia State","baseline":"84.6% enrolled in the control subgroup","result":"Enrollment increased by 3.3 percentage points","baseline_value":null,"result_value":null,"unit":null,"effect":"In the committed subgroup, enrollment rose by 3.3 percentage points from a control rate of 84.6%.","design":"Students were randomly assigned within groups defined by whether they had already committed to Georgia State.","sample":"There were 3,745 students assigned to Pounce and 3,744 controls. All had a U.S. mobile number and had agreed to receive texts.","uncertainty":"The gain applies to the committed subgroup at one university. Overall enrollment rose by 1.2 percentage points, which was not statistically significant.","comparison_basis":"measured","source_urls":["https://journals.sagepub.com/doi/10.1177/2332858417749220","https://www.wabe.org/georgia-state-uses-a-chatbot-to-attack-summer-melt/","https://edworkingpapers.com/sites/default/files/ai22-564.pdf"],"source_refs":["AIBD-0215-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The trial did not separate automated conversation from personalized reminders and counselor support.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Georgia State’s usual enrollment support.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Students were randomly assigned within groups defined by whether they had already committed to Georgia State.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"In the committed subgroup, enrollment rose by 3.3 percentage points from a control rate of 84.6%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Of students assigned to Pounce, 96.2% received texts, 85.1% replied, and 6.6% opted out.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"AIBD-0215-E01","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Researchers used university administrative records and National Student Clearinghouse enrollment data.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Enrollment, financial-aid tasks, and other requirements for entering college.","inferred":false}},"evaluations[].reproduction_materials":{"state":"not_reported","review_state":"reviewed","display":{"answer":"The article does not link to participant-level data or analysis code.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"There were 3,745 students assigned to Pounce and 3,744 controls. All had a U.S. mobile number and had agreed to receive texts.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The gain applies to the committed subgroup at one university. Overall enrollment rose by 1.2 percentage points, which was not statistically significant.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0215-reported-measure","evaluation_data":null},{"case_id":"AIBD-0296","case_url":"records/AIBD-0296.html","measure":null,"reported_result":null,"baseline":null,"result":"the device correctly identified nearly every single seizure, with a false alarm rate below that of the seizure frequency for most individuals","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://spectrum.ieee.org/this-seizuredetecting-smartwatch-could-save-your-life","https://www.accessdata.fda.gov/cdrh_docs/pdf17/K172935.pdf","https://pmc.ncbi.nlm.nih.gov/articles/PMC8418082/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13075604/"],"source_refs":["AIBD-0296-source-0001","AIBD-0296-source-0002","AIBD-0296-source-0003","AIBD-0296-source-0004","AIBD-0296-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"seizures recorded during more than 6,500 hours of monitoring of 135 patients (device test reported by IEEE Spectrum)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Result: the device correctly identified nearly every single seizure, with a false alarm rate below that of the seizure frequency for most individuals.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0296-reported-measure","evaluation_data":null},{"case_id":"AIBD-0222","case_url":"records/AIBD-0222.html","measure":"Classification accuracy, sensitivity and specificity.","reported_result":"reached 95.6% (± 4.5%) accuracy with balanced sensitivity and specificity","baseline":"Chance: 49.1% ± 13%","result":"95.6% ± 4.5%","baseline_value":null,"result_value":null,"unit":null,"effect":"95.6% (± 4.5%) accuracy against 49.1% for chance and 91.1% for random feature selection.","design":"other.","sample":"185 Swedish children (97 high-risk, 88 low-risk) from the Kronoberg reading development project, recorded at age 9 to 10.","uncertainty":"Case-control sample of extreme groups; no external validation sample; two authors hold equity in Optolexia.","comparison_basis":"measured","source_urls":["https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0165508","https://europepmc.org/article/PMC/PMC5147795","https://www.nature.com/articles/s41598-021-95275-1","https://web.archive.org/web/2026/https://schoolsweek.co.uk/struggling-readers-spotted-by-eye-tracking-software-trial/"],"source_refs":["AIBD-0222-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Chance level and random feature selection on the same cross-validation.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"95.6% (± 4.5%) accuracy against 49.1% for chance and 91.1% for random feature selection.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Support-vector machine with recursive feature elimination, scored by repeated cross-validation against high-risk and low-risk labels set from earlier reading tests.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Classification accuracy, sensitivity and specificity.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Data deposited on Figshare; an independent team reused them in 2021.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"185 Swedish children (97 high-risk, 88 low-risk) from the Kronoberg reading development project, recorded at age 9 to 10.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Case-control sample of extreme groups; no external validation sample; two authors hold equity in Optolexia.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0222-reported-measure","evaluation_data":null},{"case_id":"AIBD-0331","case_url":"records/AIBD-0331.html","measure":"ordered cancellation of the BR-153 tender budgeted at R$20 million","reported_result":"ordered cancellation of the BR-153 tender budgeted at R$20 million","baseline":"seria muito custoso porque são, em média, 200 editais por dia","result":"Os auditores do Tribunal de Contas da União recebem pontualmente às 19h um e-mail de Alice. São os resumos das centenas de contratações federais publicadas naquele dia.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://g1.globo.com/economia/tecnologia/noticia/como-as-robos-alice-sofia-e-monica-ajudam-o-tcu-a-cacar-irregularidades-em-licitacoes.ghtml","https://files.cercomp.ufg.br/weby/up/949/o/Relat%C3%B3rio_de_visita_ao_TCU_sobre_ALICE.pdf","https://web.archive.org/web/20240224093756/https://inforchannel.com.br/2017/09/27/tcu-cria-robo-para-varredura-diaria-de-irregularidades-em-editais/","https://www.jota.info/opiniao-e-analise/artigos/alice-e-a-plataforma-digital-de-controle-externo","https://europepmc.org/article/PMC/PMC10039446"],"source_refs":["AIBD-0331-source-0002","AIBD-0331-source-0003","AIBD-0331-source-0004","AIBD-0331-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: TCU auditors reading the roughly 200 federal tender notices published each day by hand","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: seria muito custoso porque são, em média, 200 editais por dia. Result: Os auditores do Tribunal de Contas da União recebem pontualmente às 19h um e-mail de Alice. São os resumos das centenas de contratações federais publicadas naquele dia.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"ordered cancellation of the BR-153 tender budgeted at R$20 million","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0331-reported-measure","evaluation_data":null},{"case_id":"AIBD-0189","case_url":"records/AIBD-0189.html","measure":"Share of inspected restaurants rated unsafe.","reported_result":"69 (52.3%) were deemed unsafe, against 2,662 of 10,786 (24.7%) baseline restaurants","baseline":"24.7% of BASELINE restaurants deemed unsafe","result":"52.3% of FINDER-identified restaurants deemed unsafe (3.06 times as likely)","baseline_value":null,"result_value":null,"unit":null,"effect":"52.3% of FINDER-prompted restaurants were unsafe against 24.7% of the others; odds ratio 3.06 (95% CI 2.14 to 4.35), adjusted for city and risk level.","design":"controlled observational.","sample":"Las Vegas (May to August 2016) and Chicago (November 2016 to March 2017); 132 FINDER-prompted inspections and 10,786 other inspections.","uncertainty":"Not randomized; small FINDER sample; Google staff co-wrote the study; no measure of illness prevented.","comparison_basis":"measured","source_urls":["https://www.nature.com/articles/s41746-018-0045-1","https://doi.org/10.1038/s41746-018-0045-1"],"source_refs":["https://www.nature.com/articles/s41746-018-0045-1"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The odds ratio compares the model's list with the existing ways of choosing restaurants to inspect.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"All other inspections in the two cities during the same period, including routine and complaint-driven ones.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"52.3% of FINDER-prompted restaurants were unsafe against 24.7% of the others; odds ratio 3.06 (95% CI 2.14 to 4.35), adjusted for city and risk level.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Daily lists of restaurants sent to the two health departments during the deployments.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Inspectors, unaware of whether FINDER prompted a visit, inspected each restaurant under the usual protocol. The authors labeled a restaurant unsafe when it received a grade that reflected a serious health code violation.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Share of inspected restaurants rated unsafe.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The search and location data are held by Google and are not public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Las Vegas (May to August 2016) and Chicago (November 2016 to March 2017); 132 FINDER-prompted inspections and 10,786 other inspections.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not randomized; small FINDER sample; Google staff co-wrote the study; no measure of illness prevented.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0189-reported-measure","evaluation_data":null},{"case_id":"AIBD-0310","case_url":"records/AIBD-0310.html","measure":"Completed outpatient appointments, new safety plans, mental health admissions, emergency department visits, documented suicide attempts, and suicide and all-cause mortality.","reported_result":"documented suicide attempts (-0.05; -0.06 to -0.03)","baseline":null,"result":"suicide attempts (ADT, −0.05; 95% CI, −0.06 to −0.03)","baseline_value":null,"result_value":null,"unit":null,"effect":"Suicide attempts ATD -0.05 (95% CI -0.06 to -0.03); mental health admissions -0.08; emergency visits -0.03; outpatient visits +0.31; new safety plans +0.08; no difference in suicide or all-cause mortality.","design":"quasi-experimental.","sample":"173,313 individuals across all VHA facilities; 40,816 in the top 0.1% tier from March 2017 to December 2018 and 36,604 in the top tier from March 2014 to December 2015.","uncertainty":"Observational design by the VA office that runs the program. Cause-of-death data covered only a subcohort, and the authors write that the mortality analyses may be underpowered. The authors also note that closer clinical attention may raise recorded attempts, so the drop in attempts may be underestimated.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8524305/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12238888/","https://europepmc.org/article/MED/36039552","https://news.va.gov/press-room/va-reach-vet-initiative-helps-save-veterans-lives-program-signals-when-more-help-is-needed-for-at-risk-veterans/","https://www.nextgov.com/artificial-intelligence/2025/07/inside-vas-yearslong-ai-effort-uncover-veterans-high-risk-suicide/406781/","https://www.nextgov.com/artificial-intelligence/2024/10/va-updating-its-ai-suicide-risk-model-reach-more-women/400377/","https://themarkup.org/news/2024/05/30/v-a-uses-a-suicide-prevention-algorithm-to-decide-who-gets-extra-help-it-favors-white-men"],"source_refs":["AIBD-0310-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The design compares program cohorts with similar risk-tier cohorts before the program. It does not separate the model's ranking from the outreach and care review that follow it.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Top 0.1% risk tier cohorts from before the program, adjusted for trends in subthreshold risk tiers.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Suicide attempts ATD -0.05 (95% CI -0.06 to -0.03); mental health admissions -0.08; emergency visits -0.03; outpatient visits +0.31; new safety plans +0.08; no difference in suicide or all-cause mortality.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Monthly identification in the top 0.1% risk tier at the veteran's facility, followed by provider outreach and care review.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Triple-difference analysis of 6-month changes in outcomes after compared with before cohort entry, using VHA health records and national death data.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Completed outpatient appointments, new safety plans, mental health admissions, emergency department visits, documented suicide attempts, and suicide and all-cause mortality.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"173,313 individuals across all VHA facilities; 40,816 in the top 0.1% tier from March 2017 to December 2018 and 36,604 in the top tier from March 2014 to December 2015.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Observational design by the VA office that runs the program. Cause-of-death data covered only a subcohort, and the authors write that the mortality analyses may be underpowered. The authors also note that closer clinical attention may raise recorded attempts, so the drop in attempts may be underestimated.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0310-reported-measure","evaluation_data":null},{"case_id":"AIBD-0219","case_url":"records/AIBD-0219.html","measure":"Share of crews who said that speed and quality of communication improved, and share who wanted to keep using the app.","reported_result":"690 of Japan's 720 fire headquarters (95.8%) used the app","baseline":null,"result":"more than 70% of crews said that both the speed and the quality of communication improved, and more than 90% wanted to keep using the app","baseline_value":null,"result_value":null,"unit":null,"effect":"More than 70% said that both speed and quality improved; more than 90% wanted to keep using the app.","design":"qualitative evaluation.","sample":"Sapporo Fire Bureau ambulance crews who used the app from January 2017; the number of respondents is not given.","uncertainty":"Self-reported views of crews in one city, reported by the developers; no count of respondents; Sapporo crews had used the app since development, as the authors note; no patient outcome.","comparison_basis":"measured","source_urls":["https://fdmc.fdma.go.jp/investigation/docs/9-12.pdf","https://www.nict.go.jp/info/topics/2017/04/170418-1.html","https://www.soumu.go.jp/johotsusintokei/whitepaper/ja/h29/html/nc155230.html","https://www.risktaisaku.com/articles/-/2758","https://www.town.ninomiya.kanagawa.jp/0000001513.html","https://www.fdma.go.jp/mission/enrichment/gaikokujin_syougaisya_torikumi/items/torikumi_shiryou2.pdf","https://www.jstage.jst.go.jp/article/jsels/5/1/5_5.1_41/_article/-char/ja/"],"source_refs":["AIBD-0219-source-0005"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not separated: the survey compares the whole app, including its 46 fixed phrases, with earlier methods.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Existing methods, such as multilingual communication boards.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"qualitative evaluation.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"More than 70% said that both speed and quality improved; more than 90% wanted to keep using the app.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Trial use from January 2017, including 29 cases in Sapporo.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"qualitative evaluation.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"A survey of ambulance crews by the Sapporo Fire Bureau after trial use, asking whether speed and quality of communication changed compared with existing methods.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Share of crews who said that speed and quality of communication improved, and share who wanted to keep using the app.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sapporo Fire Bureau ambulance crews who used the app from January 2017; the number of respondents is not given.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Self-reported views of crews in one city, reported by the developers; no count of respondents; Sapporo crews had used the app since development, as the authors note; no patient outcome.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0219-reported-measure","evaluation_data":null},{"case_id":"AIBD-0177","case_url":"records/AIBD-0177.html","measure":"False robbery reports detected and closed; share of interrogated complainants who confessed.","reported_result":"25 and 39 false robberies respectively, against 3 and 12 in the same month","baseline":"an average of 3.33 in Murcia and 12.14 in Málaga per June month","result":"25 in Murcia and 39 in Málaga in one week","baseline_value":null,"result_value":null,"unit":null,"effect":"25 and 39 in one week against monthly averages of 3.33 and 12.14; confession rates of 81.58% in Murcia and 84.78% in Málaga, which the authors give as a lower bound on precision of 83.54%.","design":"before/after.","sample":"Urban police stations in Murcia and Málaga, Spain, one week in June 2017; the number of reports checked is not given.","uncertainty":"The developers, one of them a police officer, ran the pilot. It lasted one week in two cities, and it compares one week with whole-month historical averages. The paper reports no case counts or confusion matrix. Confessions under interrogation stand in for ground truth. A satisfaction survey drew 21 participants.","comparison_basis":"measured","source_urls":["http://regulation.blogs.uv.es/files/2024/09/Three-predictive-policing-approaches-web-19.12.24.pdf","https://doi.org/10.1016/j.knosys.2018.03.010","https://www.scientificamerican.com/article/an-algorithm-that-can-spot-when-people-lie-to-the-police/","https://algorithmwatch.org/en/spain-police-veripol/","https://civio.es/transparencia/2025/03/25/national-police-stop-using-veripol-its-star-ai-for-detecting-false-reports/","https://algorithmwatch.org/en/spanish-police-halts-veripol/"],"source_refs":["AIBD-0177-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated: the experts who installed the tool also trained local officers and supervised all activity during the pilot week.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Average false robbery reports detected and closed in June of 2008 to 2016 in the same cities.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"25 and 39 in one week against monthly averages of 3.33 and 12.14; confession rates of 81.58% in Murcia and 84.78% in Málaga, which the authors give as a lower bound on precision of 83.54%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One week of use in two cities.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two expert officers installed VeriPol at police stations in each city, trained local officers and supervised its use for one week. A report counted as false only when the complainant confessed.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"False robbery reports detected and closed; share of interrogated complainants who confessed.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The paper lists the model's text features and weights; the 2015 training reports and the pilot data are not public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Urban police stations in Murcia and Málaga, Spain, one week in June 2017; the number of reports checked is not given.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The developers, one of them a police officer, ran the pilot. It lasted one week in two cities, and it compares one week with whole-month historical averages. The paper reports no case counts or confusion matrix. Confessions under interrogation stand in for ground truth. A satisfaction survey drew 21 participants.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0177-reported-measure","evaluation_data":null},{"case_id":"AIBD-0097","case_url":"records/AIBD-0097.html","measure":"reducing the area to map from 94,080 km² to 2,745 km²","reported_result":"reducing the area to map from 94,080 km² to 2,745 km²","baseline":"94,080km2","result":"2,745 km2","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.redcross.org/about-us/news-and-events/news/Data-A-vital-tool-for-disaster-response.html","https://arxiv.org/abs/1712.05839","https://www.missingmaps.org/blog/2017/08/10/malawi/","https://ai.meta.com/blog/mapping-the-world-to-help-aid-workers-with-weakly-semi-supervised-learning/"],"source_refs":["AIBD-0097-source-0001","AIBD-0097-source-0002","AIBD-0097-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"mapping the whole national land area without the settlement maps (the normal Missing Maps approach)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 94,080km2. Result: 2,745 km2.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"reducing the area to map from 94,080 km² to 2,745 km²","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0097-reported-measure","evaluation_data":null},{"case_id":"AIBD-0053","case_url":"records/AIBD-0053.html","measure":"64.0% success using the avatar at home and 70.9% with the exoskeleton in the laboratory","reported_result":"64.0% success using the avatar at home and 70.9% with the exoskeleton in the laboratory","baseline":"intracortical microelectrode systems","result":"a semi-invasive epidural system that the authors say has similar efficiency","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://europepmc.org/abstract/MED/31587955","https://europepmc.org/abstract/MED/35234665","https://www.bbc.com/news/health-49907356"],"source_refs":["AIBD-0053-source-0001","AIBD-0053-source-0002","AIBD-0053-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"64.0% success using the avatar at home and 70.9% with the exoskeleton in the laboratory","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"64.0% success using the avatar at home and 70.9% with the exoskeleton in the laboratory","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0053-reported-measure","evaluation_data":null},{"case_id":"AIBD-0621","case_url":"records/AIBD-0621.html","measure":"Sensitivity and specificity for more than mild diabetic retinopathy.","reported_result":"an observed sensitivity of 87.4%","baseline":"Goals of 85.0% sensitivity and 82.5% specificity","result":"87.4% observed sensitivity and 89.5% observed specificity","baseline_value":null,"result_value":null,"unit":null,"effect":"Sensitivity 87.4%, specificity 89.5%, imageability 96.1%.","design":"other.","sample":"900 adults with diabetes enrolled at 10 US primary care sites; 819 fully analyzable.","uncertainty":"No usual-care comparison arm; 73 participants could not be fully analyzed; vision outcomes not measured.","comparison_basis":"measured","source_urls":["https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN180001.pdf","https://pmc.ncbi.nlm.nih.gov/articles/PMC6550188"],"source_refs":["AIBD-0621-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated; the software is the whole screening test.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Reading center reference standard and pre-specified goals.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sensitivity 87.4%, specificity 89.5%, imageability 96.1%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One screening exam per participant.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Prospective diagnostic accuracy study against reading center grading of fundus photographs and OCT.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sensitivity and specificity for more than mild diabetic retinopathy.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Protocol available with the npj Digital Medicine article.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"900 adults with diabetes enrolled at 10 US primary care sites; 819 fully analyzable.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"No usual-care comparison arm; 73 participants could not be fully analyzed; vision outcomes not measured.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0621-reported-measure","evaluation_data":null},{"case_id":"AIBD-0628","case_url":"records/AIBD-0628.html","measure":"Reader AUC.","reported_result":"rose from 0.7055 to 0.7575","baseline":"AUC 0.7055","result":"AUC 0.7575","baseline_value":null,"result_value":null,"unit":null,"effect":"AUC 0.7055 to 0.7575 (proper-binormal).","design":"quasi-experimental.","sample":"19 radiologists; 111 breast MRI cases.","uncertainty":"Marginal result; trapezoidal estimate CI crossed zero; enriched cases; optimistic training material.","comparison_basis":"measured","source_urls":["https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN170022.pdf"],"source_refs":["AIBD-0628-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"AUC difference 0.0520 (95% CI 0.0022 to 0.1018).","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"First read without the score.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"AUC 0.7055 to 0.7575 (proper-binormal).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two reads per case.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sequential first read then second read with the score; DBM MRMC analysis.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Reader AUC.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"None public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"19 radiologists; 111 breast MRI cases.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Marginal result; trapezoidal estimate CI crossed zero; enriched cases; optimistic training material.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0628-reported-measure","evaluation_data":null},{"case_id":"AIBD-0308","case_url":"records/AIBD-0308.html","measure":null,"reported_result":null,"baseline":"operators check many monitors to support stations","result":"AI alerts on the monitor and by email when it detects an anomaly","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.optim.co.jp/news-detail/23884","https://www.atpress.ne.jp/news/139924","https://kabutan.jp/news/marketnews/?b=n201806270161","https://www.sankei.com/article/20141223-XETBUKCBZROQHM3IRUP6XYQ26U/"],"source_refs":["AIBD-0308-source-0001","AIBD-0308-source-0002","AIBD-0308-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0308-reported-measure","evaluation_data":null},{"case_id":"AIBD-0231","case_url":"records/AIBD-0231.html","measure":null,"reported_result":null,"baseline":"work that would be manually impossible or would take hundreds of hours","result":"seconds","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://cbn.com/news/us/creating-traffic-jam-against-sex-trafficking-groundbreaking-tech-game-changer","https://futurism.com/ai-tool-law-enforcement-stop-human-trafficking"],"source_refs":["AIBD-0231-source-0001","AIBD-0231-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0231-reported-measure","evaluation_data":null},{"case_id":"AIBD-0604","case_url":"records/AIBD-0604.html","measure":"Function scores and task times.","reported_result":"moved three clothespins in 90.2 seconds on average against 137 seconds with direct control","baseline":"137 ± 60.2 seconds for the clothespin relocation task","result":"90.2 ± 39.6 seconds for the clothespin relocation task","baseline_value":null,"result_value":null,"unit":null,"effect":"Better SHAP (p = 0.04) and faster clothespin task (p = 0.02); no difference on Box and Blocks or ACMC; seven of eight preferred pattern recognition.","design":"randomized.","sample":"Eight transhumeral amputees with TMR, Center for Bionic Medicine, Chicago.","uncertainty":"Eight participants; developers ran the trial and two hold a financial interest in Coapt; some unintended movements.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC5653840/","https://pmc.ncbi.nlm.nih.gov/articles/PMC8000641/"],"source_refs":["AIBD-0604-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"About 47 seconds faster on the clothespin task.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Direct myoelectric control.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Better SHAP (p = 0.04) and faster clothespin task (p = 0.02); no difference on Box and Blocks or ACMC; seven of eight preferred pattern recognition.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Six to eight weeks of home use under each control method.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"SHAP, Clothespin Relocation Task, Box and Blocks Test and ACMC before and after each home period; home usage logs; stated preference.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Function scores and task times.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Data available from the corresponding author on request.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Eight transhumeral amputees with TMR, Center for Bionic Medicine, Chicago.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Eight participants; developers ran the trial and two hold a financial interest in Coapt; some unintended movements.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0604-reported-measure","evaluation_data":null},{"case_id":"AIBD-0014","case_url":"records/AIBD-0014.html","measure":"a rocky planet about 30 percent larger than Earth with a 14.4-day orbit","reported_result":"a rocky planet about 30 percent larger than Earth with a 14.4-day orbit","baseline":"the weakest signals often are missed using these methods","result":"the neural network correctly identified true planets and false positives 96 percent of the time","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.nasa.gov/news-release/artificial-intelligence-nasa-data-used-to-discover-eighth-planet-circling-distant-star/","https://arxiv.org/abs/1712.05044","https://doi.org/10.3847/1538-3881/aa9e09","https://arxiv.org/abs/1903.10507","https://www.sciencenews.org/article/ai-has-found-8-planet-system-ours-kepler-data","https://exoplanetarchive.ipac.caltech.edu/overview/Kepler-90%20i","https://arxiv.org/abs/2507.13588"],"source_refs":["AIBD-0014-source-0002","AIBD-0014-source-0003","AIBD-0014-source-0001","AIBD-0014-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the automated tests and human vetting previously used to verify Kepler signals","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: the weakest signals often are missed using these methods. Result: the neural network correctly identified true planets and false positives 96 percent of the time.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a rocky planet about 30 percent larger than Earth with a 14.4-day orbit","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0014-reported-measure","evaluation_data":null},{"case_id":"AIBD-0622","case_url":"records/AIBD-0622.html","measure":"Time from CT angiogram to specialist notification.","reported_result":"a mean difference of 51.40 minutes","baseline":"58.72 minutes mean","result":"7.32 minutes mean","baseline_value":null,"result_value":null,"unit":null,"effect":"Mean 7.32 minutes against 58.72 minutes; mean difference 51.40 minutes.","design":"controlled observational.","sample":"300 archived CT angiograms; 44 true positive cases with usual-care times.","uncertainty":"Time comparison limited to the subset with usual-care times; archived scans; no patient outcomes.","comparison_basis":"measured","source_urls":["https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN170073.pdf"],"source_refs":["AIBD-0622-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean 51.40 minutes earlier alert.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Usual-care notification time recorded in radiology reports.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean 7.32 minutes against 58.72 minutes; mean difference 51.40 minutes.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One analysis per scan.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective standalone test on archived CT angiograms; device alert times compared with usual-care notification times from radiology reports.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time from CT angiogram to specialist notification.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"None public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"300 archived CT angiograms; 44 true positive cases with usual-care times.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time comparison limited to the subset with usual-care times; archived scans; no patient outcomes.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0622-reported-measure","evaluation_data":null},{"case_id":"AIBD-0314","case_url":"records/AIBD-0314.html","measure":"produced 2,930 matches among children in care","reported_result":"produced 2,930 matches among children in care","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.business-standard.com/article/pti-stories/facial-recognition-system-helps-trace-3-000-missing-children-in-4-days-118042200470_1.html","https://www.ndtv.com/delhi-news/2-percent-accuracy-in-delhi-police-facial-recognition-software-court-told-1905242","https://indiankanoon.org/doc/181795735/","https://pulitzercenter.org/stories/ai-took-over-policing-delhi-who-bore-brunt","https://pulitzercenter.org/stories/looking-back-camera-how-controversial-ai-firms-shaped-delhis-predictive-policing","https://internetfreedom.in/we-demand-the-delhi-police-stop-its-facial-recognition-system/","https://www.manoramayearbook.in/current-affairs/india/2019/07/10/Can-facial-recognition-help-fight-crime.html","https://timesofindia.indiatimes.com/city/delhi/upgrade-face-recognition-software-delhi-high-court/articleshow/70813797.cms"],"source_refs":["AIBD-0314-source-0001","AIBD-0314-source-0002","AIBD-0314-source-0003","AIBD-0314-source-0004","AIBD-0314-source-0005","AIBD-0314-source-0006","AIBD-0314-source-0007"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"matched 2,930 identities on 6–10 April 2018","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"matched 2,930 identities on 6–10 April 2018","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0314-reported-measure","evaluation_data":null},{"case_id":"AIBD-0295","case_url":"records/AIBD-0295.html","measure":"Study 1: Time below 3.0 mmol/L; participants with severe hypoglycemia. Study 2: Median percent time below 3.0 and 3.9 mmol/L; ULS use.","reported_result":"time below 54 mg/dL fell from 1.0% to 0.6%","baseline":"1.0 (1.5) percent of time <54 mg/dL on G5 (70 mg/dL low-alert setting, Table 1)","result":"0.6 (1.0) percent of time <54 mg/dL on G6 (70 mg/dL low-alert setting, Table 1)","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: Time below 3.0 mmol/L 0.47% against 0.84% (p=0.0070); severe hypoglycemia in 3 against 13 participants (p=0.0082). Study 2: Time below 3.0 mmol/L fell 0.12 to 0.28 percentage points in the first three months, with a further 0.11 to 0.21 point fall in months three to six.","design":"Study 1: randomized. Study 2: before/after.","sample":"Study 1: Six hospitals in Belgium; 254 adults with type 1 diabetes who already used FreeStyle Libre, recruited 29 January to 30 July 2019. Study 2: Users in Germany, Sweden and the United Kingdom who switched from G5 to G6 9 to 12 months after the G6 launch in each country, followed in a two-year window from 2018 to 2020.","uncertainty":"Study 1: Open-label; Dexcom funded the trial; the predictive alert cannot be separated from the other features of the G6. Study 2: No control group; country of residence taken from user-reported app data; written by Dexcom staff and a Dexcom-funded academic.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC6477579/","https://pmc.ncbi.nlm.nih.gov/articles/PMC7189147/","https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN170088.pdf","https://doi.org/10.1089/dia.2023.0376","https://doi.org/10.1089/dia.2023.0434","https://pmc.ncbi.nlm.nih.gov/articles/PMC10087512/","https://pubmed.ncbi.nlm.nih.gov/34089660/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11307228/"],"source_refs":["AIBD-0295-source-0008","AIBD-0295-source-0007"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Not isolated: the G6 arm received real-time readings and all G6 alerts together. Study 2: Not isolated; the authors could not link ULS use to outcomes.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Continued FreeStyle Libre intermittent scanning; the authors describe the switch as one to real-time CGM with alert functionality. Study 2: The same users on the G5, which had no predictive alert.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: randomized. Study 2: before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Time below 3.0 mmol/L 0.47% against 0.84% (p=0.0070); severe hypoglycemia in 3 against 13 participants (p=0.0082). Study 2: Time below 3.0 mmol/L fell 0.12 to 0.28 percentage points in the first three months, with a further 0.11 to 0.21 point fall in months three to six.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Six months of Dexcom G6 use.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: randomized. Study 2: before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Sensor glucose and reported severe hypoglycemia events over six months. Study 2: Anonymized CGM data uploaded to Dexcom Clarity, in three-month intervals over two years.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Time below 3.0 mmol/L; participants with severe hypoglycemia. Study 2: Median percent time below 3.0 and 3.9 mmol/L; ULS use.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Six hospitals in Belgium; 254 adults with type 1 diabetes who already used FreeStyle Libre, recruited 29 January to 30 July 2019. Study 2: Users in Germany, Sweden and the United Kingdom who switched from G5 to G6 9 to 12 months after the G6 launch in each country, followed in a two-year window from 2018 to 2020.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Open-label; Dexcom funded the trial; the predictive alert cannot be separated from the other features of the G6. Study 2: No control group; country of residence taken from user-reported app data; written by Dexcom staff and a Dexcom-funded academic.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0295-reported-measure","evaluation_data":null},{"case_id":"AIBD-0627","case_url":"records/AIBD-0627.html","measure":"Reader AUC, sensitivity and specificity.","reported_result":"Sensitivity rose from 74.7% to 80.3%","baseline":"Sensitivity 74.7%, specificity 88.9%","result":"Sensitivity 80.3%, specificity 91.4%","baseline_value":null,"result_value":null,"unit":null,"effect":"AUC 0.840 to 0.889; sensitivity 74.7% to 80.3%; specificity 88.9% to 91.4%.","design":"quasi-experimental.","sample":"24 clinicians of six types; 200 cases.","uncertainty":"Retrospective archived images; enriched case mix; no patient outcomes.","comparison_basis":"measured","source_urls":["https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN180005.pdf","https://pmc.ncbi.nlm.nih.gov/articles/PMC6233134"],"source_refs":["AIBD-0627-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"AUC difference 0.049 (95% CI 0.019 to 0.080).","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Unaided reads by the same clinicians.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"AUC 0.840 to 0.889; sensitivity 74.7% to 80.3%; specificity 88.9% to 91.4%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"200 aided and 200 unaided reads per clinician.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Fully crossed multi-reader multi-case study with a one-month washout; ground truth by three hand surgeons.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Reader AUC, sensitivity and specificity.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"None public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"24 clinicians of six types; 200 cases.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective archived images; enriched case mix; no patient outcomes.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0627-reported-measure","evaluation_data":null},{"case_id":"AIBD-0611","case_url":"records/AIBD-0611.html","measure":"Mean absolute error of planned case duration; wait time of following patients.","reported_result":"fell by 33.1 minutes per patient, from 49.4 to 16.3 minutes","baseline":"Mean absolute error of 59.3 minutes","result":"Mean absolute error of 49.5 minutes","baseline_value":null,"result_value":null,"unit":null,"effect":"MAE 49.5 versus 59.3 minutes; wait 16.3 versus 49.4 minutes.","design":"randomized.","sample":"683 colorectal and gynecology cases on two Memorial Sloan Kettering campuses.","uncertainty":"Single center, two services, 11 weeks, model built by the same team.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC7841577/","https://pmc.ncbi.nlm.nih.gov/articles/PMC10630142/"],"source_refs":["AIBD-0611-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"9.8 minutes less error; 33.1 minutes less waiting.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Usual EHR estimate with surgeon or scheduler input.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"MAE 49.5 versus 59.3 minutes; wait 16.3 versus 49.4 minutes.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Half of eligible cases.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Recorded patient-in-room to patient-out-of-room times against planned durations; presurgical wait times.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean absolute error of planned case duration; wait time of following patients.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Trial protocol and eMethods in supplements.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"683 colorectal and gynecology cases on two Memorial Sloan Kettering campuses.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Single center, two services, 11 weeks, model built by the same team.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0611-reported-measure","evaluation_data":null},{"case_id":"AIBD-0167","case_url":"records/AIBD-0167.html","measure":"82% (75%–89%) relative cut in eagle deaths at a Wyoming wind farm","reported_result":"82% (75%–89%) relative cut in eagle deaths at a Wyoming wind farm","baseline":"increasing at the control site by 113% (51%–218%)","result":"the number of fatalities at the treatment site declined by 63% (95% CI = 59%–66%)","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pubs.usgs.gov/publication/70219527","https://doi.org/10.1371/journal.pone.0278754","https://doi.org/10.1111/1365-2664.13831","https://doi.org/10.1111/1365-2664.14196","https://doi.org/10.1111/1365-2664.14490","https://www.rechargenews.com/wind/wind-farm-eagle-deaths-cut-by-82-as-smart-cameras-spot-birds-then-halt-turbines/2-1-953479","https://oecd.ai/en/incidents/2021-01-28-085c","https://doi.org/10.1002/2688-8319.12173","https://www.federalregister.gov/documents/2024/02/12/2024-02182/permits-for-incidental-take-of-eagles-and-eagle-nests"],"source_refs":["AIBD-0167-source-0002","AIBD-0167-source-0003","AIBD-0167-source-0004","AIBD-0167-source-0001","AIBD-0167-source-0005","AIBD-0167-source-0006","AIBD-0167-source-0007","AIBD-0167-source-0008","AIBD-0167-source-0009"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a control wind facility 15 km away without automated curtailment, before and after the installation","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: increasing at the control site by 113% (51%–218%). Result: the number of fatalities at the treatment site declined by 63% (95% CI = 59%–66%).","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"82% (75%–89%) relative cut in eagle deaths at a Wyoming wind farm","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0167-reported-measure","evaluation_data":null},{"case_id":"AIBD-0223","case_url":"records/AIBD-0223.html","measure":"by September 2021, more than 900 impostors nationally","reported_result":"by September 2021, more than 900 impostors nationally","baseline":"officials said that the passport may have passed at face value with human inspection","result":"the facial comparison reported a mismatch and the traveler was sent to secondary inspection","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.nbcnews.com/news/us-news/new-facial-recognition-tech-catches-first-impostor-d-c-airport-n903236","https://www.cbp.gov/newsroom/local-media-release/cbp-washington-dulles-international-airport-intercepted-imposter-using","https://www.cbp.gov/newsroom/local-media-release/laredo-cbp-officers-detect-impostor-through-facial-biometrics","https://www.gao.gov/assets/gao-20-568.pdf","https://gizmodo.com/cbp-facial-recognition-scanners-failed-to-find-a-single-1846251680"],"source_refs":["AIBD-0223-source-0001","AIBD-0223-source-0002","AIBD-0223-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"by September 2021, more than 900 impostors nationally","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"by September 2021, more than 900 impostors nationally","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0223-reported-measure","evaluation_data":null},{"case_id":"AIBD-0605","case_url":"records/AIBD-0605.html","measure":"Time on the ground and time until staff assistance.","reported_result":"time on the ground fell by 29.6 minutes (95% CI 20.3 to 38.9)","baseline":"Proportion of fallers on the ground for more than one hour: 31% (8 of 26)","result":"No such events in the real-time notification period","baseline_value":null,"result_value":null,"unit":null,"effect":"29.6 minutes less on the ground; 28.3 minutes less until assistance; lies over one hour fell from 31% to zero.","design":"before/after.","sample":"66 residents with dementia, 436 falls, six California memory care facilities, November 2017 to August 2018.","uncertainty":"No concurrent control; period effects possible; authors employed by or holding stock in SafelyYou.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8277400/","https://pmc.ncbi.nlm.nih.gov/articles/PMC10468707/"],"source_refs":["AIBD-0605-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"About 30 minutes less time on the ground per fall.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Baseline period without real-time notification.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"29.6 minutes less on the ground; 28.3 minutes less until assistance; lies over one hour fell from 31% to zero.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Real-time AI fall notifications with video access.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time until staff assistance and time on the ground measured automatically from video of each fall.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time on the ground and time until staff assistance.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"66 residents with dementia, 436 falls, six California memory care facilities, November 2017 to August 2018.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"No concurrent control; period effects possible; authors employed by or holding stock in SafelyYou.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0605-reported-measure","evaluation_data":null},{"case_id":"AIBD-0030","case_url":"records/AIBD-0030.html","measure":"higher sensitivity than dispatchers (84.1% versus 72.5%)","reported_result":"higher sensitivity than dispatchers (84.1% versus 72.5%)","baseline":"77.5% sensitivity for dispatchers without alerts","result":"85.0% sensitivity for the machine learning alerts alone","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://europepmc.org/article/PMC/PMC7788469","https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2774644","https://doi.org/10.1016/j.resuscitation.2019.01.015","https://eena.org/wp-content/uploads/2020_01_13_Corti_Report.pdf","https://pmc.ncbi.nlm.nih.gov/articles/PMC13665075/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12025101/","https://pubmed.ncbi.nlm.nih.gov/36634755/"],"source_refs":["AIBD-0030-source-0001","AIBD-0030-source-0002","AIBD-0030-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"higher sensitivity than dispatchers (84.1% versus 72.5%)","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"higher sensitivity than dispatchers (84.1% versus 72.5%)","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0030-reported-measure","evaluation_data":null},{"case_id":"AIBD-0322","case_url":"records/AIBD-0322.html","measure":"2.0–5.0 kg per plant against 0.5 kg","reported_result":"2.0–5.0 kg per plant against 0.5 kg","baseline":"0.5kg per Cassava plant","result":"between 2.0-5.0 kg per Cassava plant","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://plantvillage.psu.edu/blogposts/71-successful-stories-of-cassava-farmers-in-busia-county","https://bigdata.cgiar.org/inspire/inspire-challenge-2017/pest-and-disease-monitoring-by-using-artificial-intelligence/","https://pmc.ncbi.nlm.nih.gov/articles/PMC7775399/","https://www.businessinsider.com/us-state-university-developed-an-ai-assistant-for-african-farmers-2022-8"],"source_refs":["AIBD-0322-source-0001","AIBD-0322-source-0002","AIBD-0322-source-0003","AIBD-0322-source-0004","AIBD-0322-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the same farmer's harvests before she used the Nuru app","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: 0.5kg per Cassava plant. Result: between 2.0-5.0 kg per Cassava plant.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"2.0–5.0 kg per plant against 0.5 kg","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0322-reported-measure","evaluation_data":null},{"case_id":"AIBD-0018","case_url":"records/AIBD-0018.html","measure":"found 72 previously undetected pulses from the repeating fast radio burst FRB 121102","reported_result":"found 72 previously undetected pulses from the repeating fast radio burst FRB 121102","baseline":"21 bursts have been previously detected","result":"72 new pulses, totaling 93 pulses in five hours","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://arxiv.org/abs/1809.03043","https://greenbankobservatory.org/press-release/artificial-intelligence-uncovers-72-new-fast-radio-bursts-in-gbts-breakthrough-listen-data/","https://arxiv.org/html/2307.02303v2","https://arxiv.org/pdf/1910.14201"],"source_refs":["AIBD-0018-source-0001","AIBD-0018-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the conventional search algorithm previously run on the same observation","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 21 bursts have been previously detected. Result: 72 new pulses, totaling 93 pulses in five hours.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"found 72 previously undetected pulses from the repeating fast radio burst FRB 121102","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0018-reported-measure","evaluation_data":null},{"case_id":"AIBD-0651","case_url":"records/AIBD-0651.html","measure":"Study 1: Door-to-puncture time; time to specialist notification. Study 2: Time from CT angiogram to specialist notification.","reported_result":"the median time to notification was 4.31 minutes","baseline":"58.72 minutes mean","result":"7.32 minutes mean","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: Average door-to-puncture reductions of 24, 30 and 44 minutes; median notification 4.31 minutes in 4,763 patients. Study 2: 7.32 minutes against 58.72 minutes mean.","design":"Study 1: before/after. Study 2: controlled observational.","sample":"Study 1: SkyRidge (32 patients), Wellstar (120 patients), Methodist Le Bonheur (46 patients), Mount Sinai (42 patients). Study 2: Past scans from the study submitted for FDA authorization.","uncertainty":"Study 1: Small single-site cohorts; data supplied by the applicant; CMS earlier raised confounding concerns. Study 2: Time to notification is a process measure; patient outcomes were not measured in this study.","comparison_basis":"measured","source_urls":["https://www.federalregister.gov/documents/2020/09/18/2020-19637/medicare-program-hospital-inpatient-prospective-payment-systems-for-acute-care-hospitals-and-the-long-term-care-hospital","https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN170073.pdf"],"source_refs":["AIBD-0651-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Not isolated in the hospital cohorts. Study 2: About 51 minutes earlier notification.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Patients treated before ContaCT, or with occlusions that ContaCT did not flag. Study 2: Usual-care notification time.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: before/after. Study 2: controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Average door-to-puncture reductions of 24, 30 and 44 minutes; median notification 4.31 minutes in 4,763 patients. Study 2: 7.32 minutes against 58.72 minutes mean.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Stroke patients with CT angiography at hospitals using ContaCT. Study 2: One analysis per scan.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: before/after. Study 2: controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Hospital records of door-to-puncture times before and after ContaCT implementation, as reported by the applicant. Study 2: Comparison of ContaCT notification times with recorded usual-care notification times on past scans, as cited by CMS.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Door-to-puncture time; time to specialist notification. Study 2: Time from CT angiogram to specialist notification.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: None public. Study 2: None public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: SkyRidge (32 patients), Wellstar (120 patients), Methodist Le Bonheur (46 patients), Mount Sinai (42 patients). Study 2: Past scans from the study submitted for FDA authorization.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Small single-site cohorts; data supplied by the applicant; CMS earlier raised confounding concerns. Study 2: Time to notification is a process measure; patient outcomes were not measured in this study.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0651-reported-measure","evaluation_data":null},{"case_id":"AIBD-0029","case_url":"records/AIBD-0029.html","measure":"removed over 1,000 snares in December 2018, versus 200–300 in a typical month","reported_result":"removed over 1,000 snares in December 2018, versus 200–300 in a typical month","baseline":"two or three hundred snares in an average month","result":"over one thousand snares in the first month of field tests","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://arxiv.org/pdf/1903.06669","https://news.mongabay.com/2020/11/where-to-patrol-next-netflix-of-ranger-ai-serves-up-poaching-predictions/"],"source_refs":["AIBD-0029-source-0001","AIBD-0029-source-0002","AIBD-0029-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the rangers' usual monthly snare removal without PAWS","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: two or three hundred snares in an average month. Result: over one thousand snares in the first month of field tests.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"removed over 1,000 snares in December 2018, versus 200–300 in a typical month","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0029-reported-measure","evaluation_data":null},{"case_id":"AIBD-0035","case_url":"records/AIBD-0035.html","measure":"Edit time per word and editing real-time factor.","reported_result":"transcribers graded 77% of transcripts as Good","baseline":"First, a human transcriber listens to the audio and transcribes the speech.","result":"the speech transcripts are now available sooner than before","baseline_value":null,"result_value":null,"unit":null,"effect":"Edit time per word 1.52 ± 0.53 seconds for automatic drafts against 1.32 ± 0.51 for manual drafts; real-time factor 3.26 ± 1.24 against 2.66 ± 1.05.","design":"controlled observational.","sample":"239 speeches (11 hours 33 minutes) delivered at Althingi over three days and transcribed both by the contracted manual service and by the recognizer.","uncertainty":"Three days of speeches; the developers ran the study on an early version; the manual reference texts were not fully true to the audio.","comparison_basis":"assumed","source_urls":["https://www.isca-archive.org/interspeech_2019/helgadottir19_interspeech.pdf","https://www.mbl.is/frettir/innlent/2019/11/16/talgreinir_skrifar_raedur_althingismanna/","https://www.visir.is/g/2018218636d","https://catalog.ldc.upenn.edu/LDC2021S01","https://arxiv.org/pdf/1807.11893v1"],"source_refs":["https://arxiv.org/abs/1807.11893"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Negative at the time of the test: the automatic drafts added about 0.2 seconds of editing per word.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Drafts from the contracted manual transcription service for the same speeches.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Edit time per word 1.52 ± 0.53 seconds for automatic drafts against 1.32 ± 0.51 for manual drafts; real-time factor 3.26 ± 1.24 against 2.66 ± 1.05.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Editors timed their editing of each speech; word edit distance between drafts and published text; editors graded the drafts.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Edit time per word and editing real-time factor.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"239 speeches (11 hours 33 minutes) delivered at Althingi over three days and transcribed both by the contracted manual service and by the recognizer.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Three days of speeches; the developers ran the study on an early version; the manual reference texts were not fully true to the audio.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0035-reported-measure","evaluation_data":null},{"case_id":"AIBD-0052","case_url":"records/AIBD-0052.html","measure":"Correct characters per minute, selection accuracy, completion of everyday tasks, serious device-related events, vessel occlusion, and migration.","reported_result":"typed 13.81 and 20.10 correct characters per minute","baseline":"Home use began 197 days after implantation in that earlier participant","result":"Home use began on days 86 and 71 in the two Stentrode participants","baseline_value":null,"result_value":null,"unit":null,"effect":"The first two participants typed 13.81 and 20.10 correct characters per minute, with selection accuracy above 92%.","design":"A prospective, open-label feasibility study without a control group. The later report includes the first two participants.","sample":"Two implanted participants appeared in the first report. The later study enrolled five people, implanted four, and excluded one because of venous anatomy.","uncertainty":"The cohort was small and selected, and decoders differed between participants. Without a control group, superiority over other assistive systems cannot be established.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC7848062/","https://pmc.ncbi.nlm.nih.gov/articles/PMC9857731/","https://pubmed.ncbi.nlm.nih.gov/40355001/"],"source_refs":["AIBD-0052-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The study demonstrates use of the combined implant, decoder, eye tracker, and software. It does not isolate the benefit of machine learning.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"There was no concurrent control. The comparison with the Utrecht interface involved different participants and protocols.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"A prospective, open-label feasibility study without a control group. The later report includes the first two participants.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"The first two participants typed 13.81 and 20.10 correct characters per minute, with selection accuracy above 92%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"The first participants began home use without research staff on days 86 and 71 after implantation. Caregiver assistance with setup remained necessary.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"NCT03834857","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Researchers measured typing and selection accuracy, observed everyday computer tasks, recorded adverse events, and checked vessel patency and device position.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Correct characters per minute, selection accuracy, completion of everyday tasks, serious device-related events, vessel occlusion, and migration.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The 2020 supplement gives participant-specific filters, training splits, and click rules. It does not publish exact trained weights or a software release.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two implanted participants appeared in the first report. The later study enrolled five people, implanted four, and excluded one because of venous anatomy.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The cohort was small and selected, and decoders differed between participants. Without a control group, superiority over other assistive systems cannot be established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0052-reported-measure","evaluation_data":null},{"case_id":"AIBD-0107","case_url":"records/AIBD-0107.html","measure":"stage 1A melanoma was found and removed, and he is cancer-free","reported_result":"stage 1A melanoma was found and removed, and he is cancer-free","baseline":"a changing mole that he had not shown to a doctor for 12 months","result":"a stage 1A melanoma removed, with no spread found","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.dailymail.co.uk/health/article-7898507/Man-mole-removed-finding-skin-cancer-app-told-visit-doctor.html","https://www.dostor.org/2974509","https://www.skinvision.com/","https://pmc.ncbi.nlm.nih.gov/articles/PMC9367531/","https://pmc.ncbi.nlm.nih.gov/articles/PMC7190019/","https://pmc.ncbi.nlm.nih.gov/articles/PMC9393821/"],"source_refs":["AIBD-0107-source-0001","AIBD-0107-source-0002","AIBD-0107-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"stage 1A melanoma was found and removed, and he is cancer-free","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"stage 1A melanoma was found and removed, and he is cancer-free","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0107-reported-measure","evaluation_data":null},{"case_id":"AIBD-0144","case_url":"records/AIBD-0144.html","measure":"5,145 potholes were filled in North Shore for just under £450,000","reported_result":"5,145 potholes were filled in North Shore for just under £450,000","baseline":"up to £1.5 million a year","result":"£719","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.gov.uk/government/news/transport-secretary-pledges-to-fix-pothole-plague-as-she-sees-cutting-edge-repair-technology-in-action-on-blackpool-visit","https://blackpoolunlimited.com/artificial-intelligence-boosts-road-maintenance-efficiency/","https://www.blackpool.gov.uk/Residents/Parking-roads-and-transport/Roadworks-and-road-maintenance/Highways-maintenance-programme.aspx","https://www.bbc.co.uk/news/uk-england-lancashire-67892850","https://www.bbc.com/news/uk-england-lancashire-51356855","https://www.bbc.co.uk/news/articles/c4gevl5x7gdo"],"source_refs":["AIBD-0144-source-0001","AIBD-0144-source-0002","AIBD-0144-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the council's pothole compensation payouts in the years before Project Amber","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: up to £1.5 million a year. Result: £719.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"5,145 potholes were filled in North Shore for just under £450,000","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0144-reported-measure","evaluation_data":null},{"case_id":"AIBD-0615","case_url":"records/AIBD-0615.html","measure":"Test scores in standard deviations after 18 months.","reported_result":"0.22 standard deviations higher in math and 0.20 standard deviations higher in Hindi","baseline":"Comparison schools' scores (the reference level)","result":"0.22 standard deviations higher in math and 0.20 higher in Hindi","baseline_value":null,"result_value":null,"unit":null,"effect":"0.22 SD in math and 0.20 SD in Hindi.","design":"randomized.","sample":"80 Adarsh public schools (about 6,500 students) in four districts of Rajasthan, India.","uncertainty":"Working paper; no effect on school exams; Adarsh schools have better infrastructure than other public schools.","comparison_basis":"measured","source_urls":["https://www.povertyactionlab.org/evaluation/scaling-personalized-adaptive-learning-pal-math-and-language-india","https://www.nber.org/papers/w34205","https://www.nber.org/system/files/working_papers/w34205/w34205.pdf"],"source_refs":["AIBD-0615-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated; the estimate covers the software with labs, lab assistants and timetable changes.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"40 comparison schools with regular teaching.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"0.22 SD in math and 0.20 SD in Hindi.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"About a quarter of math and Hindi time in primary grades and 40 to 50 percent in middle grades over 18 months.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Independently designed and administered math and Hindi tests, scaled with item response theory.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Test scores in standard deviations after 18 months.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"AEA RCT Registry AEARCTR-0002546.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"80 Adarsh public schools (about 6,500 students) in four districts of Rajasthan, India.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Working paper; no effect on school exams; Adarsh schools have better infrastructure than other public schools.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0615-reported-measure","evaluation_data":null},{"case_id":"AIBD-0636","case_url":"records/AIBD-0636.html","measure":"Study 1: Time-weighted average of hypotension; minutes of hypotension. Study 2: Time-weighted area under the threshold of hypotension.","reported_result":"median time in hypotension fell from 32.7 to 8.0 minutes per patient","baseline":"32.7 minutes median time in hypotension per patient","result":"8.0 minutes median time in hypotension per patient","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: 0.10 against 0.44 mm Hg; 8.0 against 32.7 minutes. Study 2: Time-weighted area 0.01 against 0.08 mmHg.","design":"Study 1: randomized. Study 2: randomized.","sample":"Study 1: 68 randomized, 60 analyzed adults in elective noncardiac surgery, one tertiary center, May 2018 to March 2019. Study 2: 34 patients having esophageal or lung resection at one hospital in Zagreb, Croatia.","uncertainty":"Study 1: Small, unblinded, single center; physiological outcome only; funded by the developer. Study 2: Very small; underpowered for clinical outcomes; no outside funding.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC7078808/","https://pmc.ncbi.nlm.nih.gov/articles/PMC10061960/"],"source_refs":["AIBD-0636-source-0001","AIBD-0636-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Median difference 0.38 mm Hg and 16.7 minutes. Study 2: Fewer and shorter hypotensive episodes.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Standard care. Study 2: Conventional pulse contour goal-directed therapy.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: randomized. Study 2: randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 0.10 against 0.44 mm Hg; 8.0 against 32.7 minutes. Study 2: Time-weighted area 0.01 against 0.08 mmHg.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Patients in the intervention arm. Study 2: Patients monitored with the index.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: randomized. Study 2: randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Time-weighted average of mean arterial pressure below 65 mm Hg from continuous invasive monitoring. Study 2: Area and time-weighted area below the hypotension threshold from invasive monitoring.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Time-weighted average of hypotension; minutes of hypotension. Study 2: Time-weighted area under the threshold of hypotension.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Trial protocol in supplement. Study 2: Data on request.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 68 randomized, 60 analyzed adults in elective noncardiac surgery, one tertiary center, May 2018 to March 2019. Study 2: 34 patients having esophageal or lung resection at one hospital in Zagreb, Croatia.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Small, unblinded, single center; physiological outcome only; funded by the developer. Study 2: Very small; underpowered for clinical outcomes; no outside funding.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0636-reported-measure","evaluation_data":null},{"case_id":"AIBD-0021","case_url":"records/AIBD-0021.html","measure":"Share of patients with a new clinically actionable cardiovascular diagnosis.","reported_result":"her pulse exceeded 120 beats per minute while she was inactive","baseline":"Sinus rhythm at the stroke team review and throughout a 24-hour outpatient ECG, with no atrial fibrillation diagnosis and treatment with clopidogrel","result":"Atrial fibrillation with a fast ventricular response on ECG after the watch alarms, followed by rate control and a direct oral anticoagulant","baseline_value":null,"result_value":null,"unit":null,"effect":"30 of 264 (11.4%) received an actionable diagnosis, 6 of 41 (15%) among those with an explicit alert; number needed to evaluate about 7.","design":"other.","sample":"264 patients at Mayo Clinic sites; 41 had an explicit alert documented.","uncertainty":"Retrospective review with data extracted by one reviewer, no standard workup, short follow-up that may miss paroxysmal atrial fibrillation, and no claims data on costs.","comparison_basis":"measured","source_urls":["https://europepmc.org/article/PMC/PMC6440268","https://europepmc.org/article/PMC/PMC11083549","https://www.sciencedirect.com/science/article/abs/pii/S0735675719304747","https://academic.oup.com/eurheartj/article/41/10/1096/5651110","https://academic.oup.com/omcr/article/2019/12/495/5691267","https://europepmc.org/article/PMC/PMC7507696","https://europepmc.org/article/PMC/PMC7364102","https://europepmc.org/article/PMC/PMC7573381","https://europepmc.org/article/PMC/PMC7573479","https://europepmc.org/article/PMC/PMC8890063","https://www.dovepress.com/smart-watch-detection-of-supraventricular-tachycardia-svt-first-case-f-peer-reviewed-fulltext-article-IMCRJ","https://europepmc.org/article/PMC/PMC9269674","https://europepmc.org/article/PMC/PMC9676225","https://europepmc.org/article/PMC/PMC11462712","https://europepmc.org/article/PMC/PMC11625252","https://europepmc.org/article/PMC/PMC12412439","https://europepmc.org/article/PMC/PMC13501160","https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN180042.pdf","https://pmc.ncbi.nlm.nih.gov/articles/PMC7526465/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13118077/"],"source_refs":["AIBD-0021-source-0019"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Patients with an explicit alert against patients without one.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"30 of 264 (11.4%) received an actionable diagnosis, 6 of 41 (15%) among those with an explicit alert; number needed to evaluate about 7.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective chart review of patients whose Mayo Clinic notes mentioned an Apple Watch and an abnormal pulse.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Share of patients with a new clinically actionable cardiovascular diagnosis.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"264 patients at Mayo Clinic sites; 41 had an explicit alert documented.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective review with data extracted by one reviewer, no standard workup, short follow-up that may miss paroxysmal atrial fibrillation, and no claims data on costs.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0021-reported-measure","evaluation_data":null},{"case_id":"AIBD-0005","case_url":"records/AIBD-0005.html","measure":"sentences at a median 15.2 words per minute with a median word error rate of 25.6%","reported_result":"sentences at a median 15.2 words per minute with a median word error rate of 25.6%","baseline":"about 5 correct words or 18 correct characters per minute by typing","result":"a median 15.2 decoded words per minute with a median word error rate of 25.6%, from a 50-word vocabulary","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8972947/","https://www.advisory.com/daily-briefing/2021/07/16/paralyzed-man","https://sites.harvard.edu/sitn/2021/09/22/scientists-detect-brain-signals-to-help-a-paralyzed-man-speak","https://web.archive.org/web/20250605185915/https://www.nejm.org/doi/full/10.1056/NEJMc2113384"],"source_refs":["AIBD-0005-source-0001","AIBD-0005-source-0002","AIBD-0005-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"sentences at a median 15.2 words per minute with a median word error rate of 25.6%","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"sentences at a median 15.2 words per minute with a median word error rate of 25.6%","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0005-reported-measure","evaluation_data":null},{"case_id":"AIBD-0230","case_url":"records/AIBD-0230.html","measure":"the child was recovered within weeks","reported_result":"the child was recovered within weeks","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.wired.com/story/how-facial-recognition-fighting-child-sex-trafficking/","https://www.snopes.com/fact-check/kutcher-software-child-trafficking/"],"source_refs":["AIBD-0230-source-0001","AIBD-0230-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"claimed its tools helped identify 5,791 child victims and rescue 103","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"claimed its tools helped identify 5,791 child victims and rescue 103","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0230-reported-measure","evaluation_data":null},{"case_id":"AIBD-0625","case_url":"records/AIBD-0625.html","measure":"Share of exams with sufficient quality for four clinical assessments.","reported_result":"left ventricular size in 98.8% of exams","baseline":"Sonographer exams (reference)","result":"Agreement in qualitative visual assessment of 92.5% to 99.6% for primary parameters","baseline_value":null,"result_value":null,"unit":null,"effect":"98.8% for LV size, LV function and effusion; 92.5% for RV size.","design":"controlled observational.","sample":"Eight nurses, 240 patients, two US academic centers.","uncertainty":"No unguided nurse arm; some views reached diagnostic quality in fewer than 70% of clips.","comparison_basis":"measured","source_urls":["https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN190040.pdf","https://pmc.ncbi.nlm.nih.gov/articles/PMC8204203"],"source_refs":["AIBD-0625-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Unguided sonographer exams on the same patients.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"98.8% for LV size, LV function and effusion; 92.5% for RV size.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One guided exam per patient.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Paired exams by guided nurses and unguided sonographers, rated by five blinded cardiologists.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Share of exams with sufficient quality for four clinical assessments.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"None public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Eight nurses, 240 patients, two US academic centers.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"No unguided nurse arm; some views reached diagnostic quality in fewer than 70% of clips.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0625-reported-measure","evaluation_data":null},{"case_id":"AIBD-0356","case_url":"records/AIBD-0356.html","measure":null,"reported_result":null,"baseline":"about 200 of Sirukam's 6,000 residents worked in illegal logging","result":"a patroller says that logging has totally stopped","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.straitstimes.com/asia/se-asia/treetop-sensors-help-indonesia-eavesdrop-on-forests-to-curb-illegal-logging","https://rfcx.org/guardian","https://rfcx.org/impact","https://sumbar.antaranews.com/berita/373198/lphn-sirukam-temukan-perusak-hutan-melalui-pemasangan-alat-guardian"],"source_refs":["AIBD-0356-source-0001","AIBD-0356-source-0002","AIBD-0356-source-0003","AIBD-0356-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: logging in the area before the sensors (amid a wider enforcement crackdown)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Result: (Logging) has totally stopped.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0356-reported-measure","evaluation_data":null},{"case_id":"AIBD-0347","case_url":"records/AIBD-0347.html","measure":"the captain was convicted, the owner fined Rp300,000,000 and the vessel forfeited","reported_result":"the captain was convicted, the owner fined Rp300,000,000 and the vessel forfeited","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.interpol.int/en/News-and-Events/News/2019/INTERPOL-supports-apprehension-of-vessel-suspected-of-illegal-fishing","https://globalfishingwatch.org/article/capture-mv-nika-case/","https://oceanmind.global/","https://thepeninsulaqatar.com/article/17/07/2019/Indonesia-seizes-notorious-fishing-vessel-wanted-by-Interpol","https://www.ccamlr.org/en/compliance/non-contracting-party-iuu-vessel-list"],"source_refs":["AIBD-0347-source-0001","AIBD-0347-source-0002","AIBD-0347-source-0003","AIBD-0347-source-0004","AIBD-0347-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the captain was convicted, the owner fined Rp300,000,000 and the vessel forfeited","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the captain was convicted, the owner fined Rp300,000,000 and the vessel forfeited","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0347-reported-measure","evaluation_data":null},{"case_id":"AIBD-0175","case_url":"records/AIBD-0175.html","measure":"Approval of low-credit-score applicants, offered rates, later financial outcomes.","reported_result":"approves 15 to 30 percent of low-score applicants rejected by traditional models, at lower rates","baseline":"approval rate and average APRs of the traditional model (not quantified)","result":"approves 27% more applicants than the traditional model, and yields 16% lower average APRs for approved loans","baseline_value":null,"result_value":null,"unit":null,"effect":"The platform approves 15 to 30 percent of low-credit-score applicants that traditional models reject and offers them lower rates.","design":"quasi-experimental.","sample":"Applicants to Upstart, a U.S. fintech lender of unsecured personal loans.","uncertainty":"Data from one lender, supplied by that lender.","comparison_basis":"measured","source_urls":["https://www.consumerfinance.gov/about-us/blog/update-credit-access-and-no-action-letter/","https://www.consumerfinance.gov/about-us/newsroom/cfpb-issues-order-to-terminate-upstart-no-action-letter/","https://www.relmanlaw.com/media/cases/1511_Upstart%20Final%20Report.pdf","https://web.archive.org/web/20260515125413/https://pubsonline.informs.org/doi/10.1287/mnsc.2024.07854","https://protectborrowers.org/wp-content/uploads/2020/02/Education-Redlining-Report.pdf"],"source_refs":["AIBD-0175-source-0004"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Alternative data drives most gains for thin-file borrowers; model sophistication adds gains where credit files are fuller.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Traditional underwriting models without alternative data.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"The platform approves 15 to 30 percent of low-credit-score applicants that traditional models reject and offers them lower rates.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Anonymized lending data from Upstart, with approvals compared with those of traditional models; later outcomes estimated from exogenous variation in credit access.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Approval of low-credit-score applicants, offered rates, later financial outcomes.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Applicants to Upstart, a U.S. fintech lender of unsecured personal loans.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Data from one lender, supplied by that lender.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0175-reported-measure","evaluation_data":null},{"case_id":"AIBD-0352","case_url":"records/AIBD-0352.html","measure":"diagnostic sensitivity and specificity for parasite detection; species identification; parasite density within 25% of reference.","reported_result":"Diagnostic sensitivity was 91.1% (95% CI 88.9–92.7) and specificity 75.6% (73.1–78.0)","baseline":null,"result":"diagnostic sensitivity of the EasyScan Go device was 91.1% (95% CI 88.9–92.7) and specificity was 75.6% (95% CI 73.1–78.0)","baseline_value":null,"result_value":null,"unit":null,"effect":"sensitivity 91.1% (95% CI 88.9 to 92.7) and specificity 75.6% (95% CI 73.1 to 78.0); species correct in 93% of P. falciparum and 92% of P. vivax samples; density within 25% in 23% of slides.","design":"controlled observational.","sample":"2,250 blood films at 11 research sites in 11 countries in Africa, Asia and South America, June 2018 to October 2019.","uncertainty":"All measures depended on slide quality; specificity rose to 85.1% on good slides; sensitivity fell to 57% below about 200 parasites per microliter; no routine field use and no patient outcomes; maker-funded with six maker employees among the authors.","comparison_basis":"measured","source_urls":["https://www.ebi.ac.uk/europepmc/webservices/rest/PMC9004086/fullTextXML","https://europepmc.org/article/PMC/PMC9004086","https://doi.org/10.1186/s12936-022-04146-1","https://www.frontiersin.org/articles/10.3389/fmala.2023.1148115/full"],"source_refs":["AIBD-0352-source-0004"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The device's readings are the AI output, so the figures measure the AI's own accuracy against experts; no human-plus-AI arm was tested.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"expert light microscopy.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"sensitivity 91.1% (95% CI 88.9 to 92.7) and specificity 75.6% (95% CI 73.1 to 78.0); species correct in 93% of P. falciparum and 92% of P. vivax samples; density within 25% in 23% of slides.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Each slide read by the device and by two expert microscopists, with repeat reads on the same or a second device to test reliability; staff trained on device operation at each site at the start of the study.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Giemsa-stained blood films read independently by expert microscopists and by the EasyScan GO; STARD reporting; ClinicalTrials.gov NCT03512678.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"diagnostic sensitivity and specificity for parasite detection; species identification; parasite density within 25% of reference.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Protocol registered as NCT03512678; software version 1.15.17b.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"2,250 blood films at 11 research sites in 11 countries in Africa, Asia and South America, June 2018 to October 2019.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"All measures depended on slide quality; specificity rose to 85.1% on good slides; sensitivity fell to 57% below about 200 parasites per microliter; no routine field use and no patient outcomes; maker-funded with six maker employees among the authors.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0352-reported-measure","evaluation_data":null},{"case_id":"AIBD-0614","case_url":"records/AIBD-0614.html","measure":"Full ENEM essay score in standard deviations.","reported_result":"raised essay scores by 0.094 standard deviations","baseline":"Comparison schools' ENEM essay scores (the reference level)","result":"0.094 standard deviations higher in schools with either version of the system","baseline_value":null,"result_value":null,"unit":null,"effect":"0.094 standard deviations for both the automated-only and the automated-plus-human-grader versions.","design":"randomized.","sample":"178 public high schools in Espírito Santo, Brazil, with about 19,000 senior students; 55 schools per treatment arm and 68 comparison schools.","uncertainty":"Working paper; pooled outcome across two essays; the company is not named in the paper.","comparison_basis":"measured","source_urls":["https://osf.io/qw249/download","https://www.povertyactionlab.org/evaluation/impact-artificial-intelligence-learning-brazil","https://exame.com/brasil/letrus-fara-estudo-com-40-mil-alunos-para-avaliar-impacto-da-ia-para-melhorar-escrita/"],"source_refs":["AIBD-0614-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The automated-only arm matched the arm with human graders, so the gain does not depend on human grading.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Comparison schools without the platform.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"0.094 standard deviations for both the automated-only and the automated-plus-human-grader versions.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Five practice essays over the 2019 school year.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"ENEM essay scores from the official 2019 exam and an independently administered essay with the same structure, graded on the ENEM's five competencies.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Full ENEM essay score in standard deviations.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"AEA RCT Registry AEARCTR-0003729.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"178 public high schools in Espírito Santo, Brazil, with about 19,000 senior students; 55 schools per treatment arm and 68 comparison schools.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Working paper; pooled outcome across two essays; the company is not named in the paper.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0614-reported-measure","evaluation_data":null},{"case_id":"AIBD-0623","case_url":"records/AIBD-0623.html","measure":"Adenoma detection rate and adenomas per colonoscopy.","reported_result":"an adenoma detection rate of 55.1% with GI Genius against 42.0% with standard colonoscopy","baseline":"42.0% adenoma detection rate","result":"55.1% adenoma detection rate","baseline_value":null,"result_value":null,"unit":null,"effect":"ADR 55.1% against 42.0%; APC 0.81 against 0.57.","design":"randomized.","sample":"700 patients at three Italian hospitals; 263 low-risk patients in the FDA analysis; six endoscopists.","uncertainty":"Limited patient diversity; six endoscopists in Italy; cancer outcomes not measured.","comparison_basis":"measured","source_urls":["https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN200055.pdf","https://pubmed.ncbi.nlm.nih.gov/32371116"],"source_refs":["AIBD-0623-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Adjusted ADR difference of 13.1 percentage points.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Standard colonoscopy without the program.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"ADR 55.1% against 42.0%; APC 0.81 against 0.57.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One colonoscopy per patient.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Randomized 1:1 multicenter trial with histology as reference; FDA reanalysis in low-risk patients.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Adenoma detection rate and adenomas per colonoscopy.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"ClinicalTrials.gov NCT04079478.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"700 patients at three Italian hospitals; 263 low-risk patients in the FDA analysis; six endoscopists.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Limited patient diversity; six endoscopists in Italy; cancer outcomes not measured.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0623-reported-measure","evaluation_data":null},{"case_id":"AIBD-0634","case_url":"records/AIBD-0634.html","measure":"Time to TB treatment; proportion treated the same day.","reported_result":"Median time to treatment was 1 day in the X-ray arm and 11 days under standard care","baseline":"0 of 5 TB patients treated on the day of the visit; median 11 days to treatment","result":"6 of 15 TB patients treated on the day of the visit; median 1 day to treatment","baseline_value":null,"result_value":null,"unit":null,"effect":"Median 1 day to treatment against 11 days under standard care; hazard ratio 2.86 (1.04 to 7.87); same-day treatment 6/15 against 0/5.","design":"randomized.","sample":"Bangwe Health Centre, Blantyre, Malawi; 1,462 adults with cough randomized 1:1:1 between 15 November 2018 and 27 November 2019.","uncertainty":"Few TB cases (15, 8 and 5 by arm), a difference in the number treated that was not statistically significant, 56-day follow-up, an open-label design, and a 0% probability of cost-effectiveness at US$1,200 per QALY.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8459969/"],"source_refs":["AIBD-0634-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated; the X-ray and software were delivered as one package.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Standard care; HIV testing alone.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Median 1 day to treatment against 11 days under standard care; hazard ratio 2.86 (1.04 to 7.87); same-day treatment 6/15 against 0/5.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"497 adults allocated to HIV testing plus digital X-ray with CAD4TB v5.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time to TB treatment and same-day treatment from clinic and study records over 56 days; intention-to-treat analysis.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time to TB treatment; proportion treated the same day.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Trial registered as NCT03519425.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Bangwe Health Centre, Blantyre, Malawi; 1,462 adults with cough randomized 1:1:1 between 15 November 2018 and 27 November 2019.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Few TB cases (15, 8 and 5 by arm), a difference in the number treated that was not statistically significant, 56-day follow-up, an open-label design, and a 0% probability of cost-effectiveness at US$1,200 per QALY.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0634-reported-measure","evaluation_data":null},{"case_id":"AIBD-0672","case_url":"records/AIBD-0672.html","measure":"Log cruising time.","reported_result":"cut taxi drivers' cruising time by 5% on average and 7% for low-skilled drivers","baseline":"Cruising time without the app","result":"5% shorter on average; 7% shorter for low-skilled drivers; skill gap narrowed by 13.4%","baseline_value":null,"result_value":null,"unit":null,"effect":"5% shorter on average, 7% for low-skilled drivers, near zero for high-skilled drivers.","design":"controlled observational.","sample":"About 500 Yokohama taxi drivers over 29 days.","uncertainty":"Drivers chose when to use the app; one city; one month; developer not named.","comparison_basis":"measured","source_urls":["https://www.nber.org/system/files/working_papers/w30612/w30612.pdf","https://doi.org/10.1287/mnsc.2023.01631"],"source_refs":["AIBD-0672-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Productivity gap narrowed by 13.4% (published) or 14% (working paper).","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Same drivers with the app off, with driver, ward and date-hour fixed effects.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"5% shorter on average, 7% for low-skilled drivers, near zero for high-skilled drivers.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"App turned on while cruising.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Cruising time until the next customer, from app and driving records.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Log cruising time.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Data from the company; not public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"About 500 Yokohama taxi drivers over 29 days.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Drivers chose when to use the app; one city; one month; developer not named.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0672-reported-measure","evaluation_data":null},{"case_id":"AIBD-0191","case_url":"records/AIBD-0191.html","measure":"Diagnostic accuracy","reported_result":"Mean accuracy was 65% for Nuru in 2020 and 49% for trained extension officers; the authors report Tukey HSD p < 0.05.","baseline":"Trained agricultural extension officers","result":"Nuru, June 2020","baseline_value":49,"result_value":65,"unit":"percent","effect":"The 2020 app averaged 65% correct diagnoses, compared with 49% for trained extension officers.","design":"Comparative diagnostic accuracy study.","sample":"The CaSRAT comparison used 170 images from Tanzania.","uncertainty":"This compares diagnostic accuracy on displayed images, not management decisions or crop yields. Human and device observations have different units of replication.","comparison_basis":"measured","source_urls":["https://doi.org/10.3389/fpls.2020.590889"],"source_refs":["AIBD-0191-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Nuru exceeded the trained extension officers by 16 percentage points on the image benchmark. This does not establish better field decisions or higher crop yields.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The 2018 app and human assessors.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Comparative diagnostic accuracy study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"The 2020 app averaged 65% correct diagnoses, compared with 49% for trained extension officers.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"170 images in the CaSRAT comparison.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"The CaSRAT comparison reported by Mrisho and colleagues in 2020.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"The percentage of leaf images that matched expert diagnoses.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Diagnostic accuracy on the CaSRAT image set.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed","display":{"answer":"The article links supplementary material; its completeness for replication is unconfirmed.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"The CaSRAT comparison used 170 images from Tanzania.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Performance varied by symptom severity. The 25-point increase between app versions used different phone sets, so it does not isolate a software improvement. Table 4 reports spreads of ±10% and ±3% without identifying their type.","inferred":false}}},"extraction_basis":"Original methods and results inspected on October 28, 2026.","observation_id":"AIBD-0191-CaSRAT-2020-trained-extension-officers","baseline_sample_size":10,"baseline_sample_unit":"people","effect_value":16,"effect_unit":"percentage points","effect_calculation":"65 - 49","sample_size":170,"sample_unit":"images","source_locator":"Methods, CaSRAT comparison; Results, Figure 3 and paragraph before Table 4"},{"case_id":"AIBD-0191","case_url":"records/AIBD-0191.html","measure":"Diagnostic accuracy","reported_result":"2020 mean: 65% ± 3%; 2018 mean: 40% ± 10%. The versions were tested on different phone sets.","baseline":"Nuru, March 2018","result":"Nuru, June 2020","baseline_value":40,"result_value":65,"unit":"percent","effect":"The 2020 app averaged 65% correct diagnoses, compared with 49% for trained extension officers.","design":"Comparative diagnostic accuracy study.","sample":"The CaSRAT comparison used 170 images from Tanzania.","uncertainty":"Performance varied by symptom severity. The 25-point increase between app versions used different phone sets, so it does not isolate a software improvement. Table 4 reports spreads of ±10% and ±3% without identifying their type.","comparison_basis":"measured","source_urls":["https://doi.org/10.3389/fpls.2020.590889"],"source_refs":["AIBD-0191-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Nuru exceeded the trained extension officers by 16 percentage points on the image benchmark. This does not establish better field decisions or higher crop yields.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The 2018 app and human assessors.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Comparative diagnostic accuracy study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"The 2020 app averaged 65% correct diagnoses, compared with 49% for trained extension officers.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"170 images in the CaSRAT comparison.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"The CaSRAT comparison reported by Mrisho and colleagues in 2020.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"The percentage of leaf images that matched expert diagnoses.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Diagnostic accuracy on the CaSRAT image set.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed","display":{"answer":"The article links supplementary material; its completeness for replication is unconfirmed.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"The CaSRAT comparison used 170 images from Tanzania.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Performance varied by symptom severity. The 25-point increase between app versions used different phone sets, so it does not isolate a software improvement. Table 4 reports spreads of ±10% and ±3% without identifying their type.","inferred":false}}},"extraction_basis":"Original Table 4 inspected on October 28, 2026.","observation_id":"AIBD-0191-CaSRAT-2020-app-versions","baseline_reported_spread":10,"result_reported_spread":3,"spread_unit":"percentage points","spread_type":"Unspecified in Table 4","effect_value":25,"effect_unit":"percentage points","effect_calculation":"65 - 40","sample_size":170,"sample_unit":"images","source_locator":"Table 4"},{"case_id":"AIBD-0337","case_url":"records/AIBD-0337.html","measure":"Likelihood that a taxpayer reviews the return and adjusts the data.","reported_result":"adjustments estimated to have a revenue impact of around $37 million","baseline":null,"result":"2.5 times more likely to review their return and 3 times more likely to adjust their data","baseline_value":null,"result_value":null,"unit":null,"effect":"2.5 times more likely to review the return and 3 times more likely to adjust the data.","design":"randomized.","sample":"Self-lodging myTax users, 2023-24 income year; sample size not stated.","uncertainty":"The ATO ran and reported the trial itself; the APSC summary gives no sample size, confidence interval or effect on revenue from the trial.","comparison_basis":"measured","source_urls":["https://web.archive.org/web/2026/https://www.apsc.gov.au/initiatives-and-programs/workforce-information/research-analysis-and-publications/state-service/state-service-report-2024-25/ways-working-ai-aps/nudge-messaging-using-real-time-analytics-tax-time","https://www.ato.gov.au/about-ato/commitments-and-reporting/information-and-privacy/how-we-use-data-and-analytics","https://web.archive.org/web/20210714025448/https://www.ato.gov.au/About-ATO/Commitments-and-reporting/Information-and-privacy/How-we-use-data-and-analytics/","https://www.themandarin.com.au/287193-auditor-whacks-ato-over-artificial-intelligence/","https://web.archive.org/web/2026/https://www.anao.gov.au/work/performance-audit/governance-of-artificial-intelligence-the-australian-taxation-office"],"source_refs":["AIBD-0337-source-0008"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated: the trial tests the message, and the models choose which claims are flagged.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Taxpayers who did not receive a nudge message.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"2.5 times more likely to review the return and 3 times more likely to adjust the data.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Over 712,000 prompts in the 2023-24 income year.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"ATO randomized control trial of real-time nudge messages in myTax, as summarized by the APSC.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Likelihood that a taxpayer reviews the return and adjusts the data.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Self-lodging myTax users, 2023-24 income year; sample size not stated.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The ATO ran and reported the trial itself; the APSC summary gives no sample size, confidence interval or effect on revenue from the trial.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0337-reported-measure","evaluation_data":null},{"case_id":"AIBD-0284","case_url":"records/AIBD-0284.html","measure":"a sudden drop in heart rate from 99 to 65 beats per minute at 19:29","reported_result":"a sudden drop in heart rate from 99 to 65 beats per minute at 19:29","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC7035811/","https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN180042.pdf"],"source_refs":["AIBD-0284-source-0001","AIBD-0284-source-0002","AIBD-0284-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a concurrent 24-hour Holter monitor","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"sudden 99-to-65 bpm drop at 19:29 that coincided with reversion from AF to sinus rhythm","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"sudden 99-to-65 bpm drop at 19:29 that coincided with reversion from AF to sinus rhythm","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0284-reported-measure","evaluation_data":null},{"case_id":"AIBD-0108","case_url":"records/AIBD-0108.html","measure":"eight occupants walked away","reported_result":"eight occupants walked away","baseline":"the driver says that he could not have reacted in time and that a further second would have put the car under the tree","result":"the car stopped with the trunk on the bonnet and all occupants survived","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.mirror.co.uk/news/uk-news/tesla-autopilot-saves-familys-life-21512255","https://www.bournemouthecho.co.uk/news/18245292.tesla-drivers-lucky-escape-falling-tree-hits-cars-storm-dennis/","https://www.bournemouthecho.co.uk/news/18240206.miracle-falling-tree-narrowly-misses-killing-families-two-tesla-model-xs---cars-automatic-emergency-brakes/","https://www.thesun.co.uk/motors/10979882/families-miracle-escape-teslas-automatic-brakes-stopped-tree/","https://twitter.com/sterlingandlaw/status/1228819929832525826","https://www.mirror.co.uk/news/uk-news/elon-musk-glad-everyone-ok-21527013","https://www.lavanguardia.com/motor/actualidad/20200219/473657698342/el-autopilot-de-tesla-salva-la-vida-a-8-personas-tormenta-dennis-reino-unido-arbol-elon-musk.html","https://www.rac.co.uk/drive/news/autonomous-vehicles-news/tesla-model-xs-automatic-brakes-just-saved-two-families-from-a-falling-tree/","https://electrek.co/2020/02/17/tesla-autopilot-saved-family-freak-accident-giant-tree-fell-model-x-suvs/","https://www.mirror.co.uk/news/uk-news/toddler-hijacks-mums-bbc-interview-21578887"],"source_refs":["AIBD-0108-source-0002","AIBD-0108-source-0003","AIBD-0108-source-0001","AIBD-0108-source-0004","AIBD-0108-source-0005","AIBD-0108-source-0006","AIBD-0108-source-0007","AIBD-0108-source-0008"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"eight occupants walked away","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"eight occupants walked away","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0108-reported-measure","evaluation_data":null},{"case_id":"AIBD-0016","case_url":"records/AIBD-0016.html","measure":null,"reported_result":null,"baseline":"vehicle or the antibiotic metronidazole (lower clearance rate)","result":"sterilization of 3 out of 4 mice after 72 hr of treatment, and 4 out of 4 mice after 96 hr of treatment","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8349178/","https://pmc.ncbi.nlm.nih.gov/articles/PMC8698312/","https://academic.oup.com/femspd/article/80/1/ftac037/6713948","https://pmc.ncbi.nlm.nih.gov/articles/PMC11111955/","https://www.news-medical.net/health/Halicin-The-AI-Discovered-Antibiotic-That-Fights-Superbugs.aspx"],"source_refs":["AIBD-0016-source-0001","AIBD-0016-source-0002","AIBD-0016-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"vehicle control and the first-line antibiotic metronidazole in a mouse Clostridioides difficile infection model","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: vehicle or the antibiotic metronidazole (lower clearance rate). Result: sterilization of 3 out of 4 mice after 72 hr of treatment, and 4 out of 4 mice after 96 hr of treatment.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0016-reported-measure","evaluation_data":null},{"case_id":"AIBD-0193","case_url":"records/AIBD-0193.html","measure":"protected the food security of 36.6 million people and avoided USD 1.56 billion in losses","reported_result":"protected the food security of 36.6 million people and avoided USD 1.56 billion in losses","baseline":"loss of 4 million tonnes of cereal and 800 million litres of milk production, USD 1.56 billion in cereal and milk losses","result":"those losses prevented and the food security of 36.6 million people protected","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.fao.org/newsroom/story/A-high-tech-response-is-helping-countries-win-the-battle-against-Desert-Locusts/en","https://www.nytimes.com/2021/04/08/science/locust-swarms-africa.html","https://www.psu.edu/news/research/story/penn-state-responds-app-aids-un-efforts-control-africas-locust-infestation","https://www.fao.org/newsroom/detail/fao-desert-locust-campaign-east-africa-upsurge-supressed-kenya-free-from-desert-locust/en","https://pmc.ncbi.nlm.nih.gov/articles/PMC13251092/"],"source_refs":["AIBD-0193-source-0001","AIBD-0193-source-0002","AIBD-0193-source-0003","AIBD-0193-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: no locust control operations in East Africa (FAO's estimate of the losses that the control campaign averted)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: loss of 4 million tonnes of cereal and 800 million litres of milk production, USD 1.56 billion in cereal and milk losses. Result: those losses prevented and the food security of 36.6 million people protected.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"protected the food security of 36.6 million people and avoided USD 1.56 billion in losses","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0193-reported-measure","evaluation_data":null},{"case_id":"AIBD-0109","case_url":"records/AIBD-0109.html","measure":null,"reported_result":null,"baseline":"Peters had not yet seen the oncoming car or touched the brake","result":"The car had braked and begun a lane change to the right","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://web.archive.org/web/20200828060021/https://www.fox46.com/news/mans-tesla-swerves-away-from-driver-avoiding-head-on-collision/","https://www.dailymail.co.uk/news/article-8092899/Model-S-Tesla-swerves-Autopilot-avoid-wrong-way-driver-dark-North-Carolina-highway.html","https://electrek.co/2020/03/09/tesla-autopilot-automatically-avoids-head-on-collision-car-wrong-lane/","https://cleantechnica.com/2020/03/08/dear-tesla-elon-musk-because-of-you-i-wont-be-going-to-a-funeral/"],"source_refs":["AIBD-0109-source-0001","AIBD-0109-source-0002","AIBD-0109-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0109-reported-measure","evaluation_data":null},{"case_id":"AIBD-0063","case_url":"records/AIBD-0063.html","measure":"Number of incidental resectable cancers flagged by the software, stage at resection, and timing of pulmonology consultation relative to the radiologist's report.","reported_result":"13 (17.3%) had cancers found incidentally because the software flagged a nodule","baseline":"Radiologists' reports described a lung abnormality in 12 of 13 patients and read no active lung disease in one","result":"The software flagged a nodule in all 13 patients","baseline_value":null,"result_value":null,"unit":null,"effect":"All 13 nodules were flagged; 8 patients saw a pulmonologist on the day of the radiograph before the official report; 7 were stage I.","design":"case study.","sample":"75 patients with surgically resected lung cancer at Yongin Severance Hospital from March 2020 to February 2022, of whom 13 had cancers found incidentally on radiographs taken for other reasons.","uncertainty":"Small single-hospital case series with no control group and no survival data; clinicians may have acted on the radiologist's report.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10004566/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11335519/"],"source_refs":["AIBD-0063-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not measured; the authors assume from timestamps that the AI results prompted the same-day consultations.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"None; the authors did not compare consultations or outcomes with a period or group without AI.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"All 13 nodules were flagged; 8 patients saw a pulmonologist on the day of the radiograph before the official report; 7 were stage I.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"AI analysis of every chest radiograph, with results shown in the image archive immediately.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective review of medical records, radiology reports, AI abnormality scores, consultation timestamps and pathology.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Number of incidental resectable cancers flagged by the software, stage at resection, and timing of pulmonology consultation relative to the radiologist's report.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"75 patients with surgically resected lung cancer at Yongin Severance Hospital from March 2020 to February 2022, of whom 13 had cancers found incidentally on radiographs taken for other reasons.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small single-hospital case series with no control group and no survival data; clinicians may have acted on the radiologist's report.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0063-reported-measure","evaluation_data":null},{"case_id":"AIBD-0250","case_url":"records/AIBD-0250.html","measure":"Time before a controller sent a risk warning to a vessel; number of warnings issued to risk-prone vessels.","reported_result":"the number of warnings issued to risk-prone vessels nearly doubled","baseline":"conventional operation (no absolute figure given)","result":"risk warning sent an average of approximately two minutes earlier; number of warnings issued for risk-prone vessels nearly doubled","baseline_value":null,"result_value":null,"unit":null,"effect":"Warnings about two minutes earlier on average; the number of warnings to risk-prone vessels nearly doubled.","design":"simulation.","sample":"Six experienced controllers and instructors in the training environment of the Tokyo Wan Vessel Traffic Service Center, December 2019 to March 2020, with 36 analysis patterns across monitoring areas and controller experience.","uncertainty":"Six participants, replayed cases and no live traffic. Fujitsu, the developer, reports no variance or significance figures. Fujitsu's 2021 release says that conventional collision-risk prediction, which assumes straight-line travel, raised many unnecessary alerts on curved routes, so some of the added warnings may have been unneeded.","comparison_basis":"measured","source_urls":["https://www.maritime-executive.com/article/ai-system-tested-by-tokyo-bay-vessel-traffic-controllers","https://oecd.ai/en/incidents/2020-04-16-d2f1","https://info.archives.global.fujitsu/global/about/resources/news/press-releases/2021/0928-01.html","https://info.archives.global.fujitsu/global/about/resources/news/press-releases/2020/0415-01.html"],"source_refs":["AIBD-0250-source-0004"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Conventional control operation without the system's risk information.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"simulation.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Warnings about two minutes earlier on average; the number of warnings to risk-prone vessels nearly doubled.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Simulated sessions; no hours or vessel counts reported.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"simulation.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Controllers handled simulated control tasks built from past near-miss cases, once with conventional methods and once with risk information computed from past AIS data and Coast Guard risk-recommendation records; Fujitsu reports a statistical analysis of the two conditions.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time before a controller sent a risk warning to a vessel; number of warnings issued to risk-prone vessels.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Six experienced controllers and instructors in the training environment of the Tokyo Wan Vessel Traffic Service Center, December 2019 to March 2020, with 36 analysis patterns across monitoring areas and controller experience.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Six participants, replayed cases and no live traffic. Fujitsu, the developer, reports no variance or significance figures. Fujitsu's 2021 release says that conventional collision-risk prediction, which assumes straight-line travel, raised many unnecessary alerts on curved routes, so some of the added warnings may have been unneeded.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0250-reported-measure","evaluation_data":null},{"case_id":"AIBD-0275","case_url":"records/AIBD-0275.html","measure":"For vision-threatening retinopathy the system's accuracy was 94.7%","reported_result":"For vision-threatening retinopathy the system's accuracy was 94.7%","baseline":"an accuracy of 93·5 (91·7-95·0; p=0·17), a sensitivity of 84·8% (79·4-90·0; p=0·024), and specificity of 95·5% (94·1-96·7; p=0·98)","result":"an accuracy of 94·7% (95% CI 93·0-96·2), sensitivity of 91·4% (87·1-95·0), and specificity of 95·4% (94·1-96·7)","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=TITLE:%22Real-time%20diabetic%20retinopathy%20screening%20by%20deep%20learning%20in%20a%20multisite%20national%20screening%20programme%22&resultType=core&format=json","https://doi.org/10.1016/s2589-7500(22)00017-6","https://www.technologyreview.com/2020/04/27/1000658/google-medical-ai-accurate-lab-real-life-clinic-covid-diabetes-retina-disease/","https://www.nationthailand.com/business/tech/40046121","https://www.nature.com/articles/s41591-026-04643-9","https://aiweekly.co/alerts/google-arda-eye-ai-tops-a-million-patients-in-three-countries","https://pmc.ncbi.nlm.nih.gov/articles/PMC11754548/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13272687/"],"source_refs":["AIBD-0275-source-0001","AIBD-0275-source-0002","AIBD-0275-source-0003","AIBD-0275-source-0004","AIBD-0275-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"regional retina specialist over-readers, measured against an adjudicated reference standard from fellowship-trained retina specialists","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: an accuracy of 93·5 (91·7-95·0; p=0·17), a sensitivity of 84·8% (79·4-90·0; p=0·024), and specificity of 95·5% (94·1-96·7; p=0·98). Result: an accuracy of 94·7% (95% CI 93·0-96·2), sensitivity of 91·4% (87·1-95·0), and specificity of 95·4% (94·1-96·7).","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"94.7% accuracy for vision-threatening DR, 2,412 referrals","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0275-reported-measure","evaluation_data":null},{"case_id":"AIBD-0307","case_url":"records/AIBD-0307.html","measure":null,"reported_result":null,"baseline":"no automatic alert to the station office when a person fell or lingered at the platform edge","result":"an alert with camera video on a terminal in the station office, so that staff can stop trains","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://connect.panasonic.com/jp-ja/case-studies/odakyu","https://www.watch.impress.co.jp/docs/news/1243957.html","https://www.nikkei.com/article/DGXLRSP531980_Q0A330C2000000/","https://www.watch.impress.co.jp/docs/news/1629496.html","https://www.odakyu.jp/sustainability/safety/safety_report/webook/2025/book/pdf/all.pdf","https://www.odakyu.jp/sustainability/safety/safety_report/webook/2026/book/pdf/all.pdf"],"source_refs":["AIBD-0307-source-0001","AIBD-0307-source-0002","AIBD-0307-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0307-reported-measure","evaluation_data":null},{"case_id":"AIBD-0420","case_url":"records/AIBD-0420.html","measure":null,"reported_result":null,"baseline":"传统方法 (traditional methods)","result":"Images flagged by the tool for review and reporting","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://mp.weixin.qq.com/s/KFJ2Mf6HA2ajyTQu7t-HpA","https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-1-PDF-E.pdf","https://www.ifaw.org/press-releases/disrupting-wildlife-trade-advanced-ai-solution","https://www.ifaw.org/press-releases/ai-endangered-species-recognize-images-illegal-wildlife"],"source_refs":["AIBD-0420-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0420-reported-measure","evaluation_data":null},{"case_id":"AIBD-0200","case_url":"records/AIBD-0200.html","measure":"increased commercially sellable production by 6% and net profit by 17%","reported_result":"increased commercially sellable production by 6% and net profit by 17%","baseline":"the growers' sellable yield and net profit in 2018","result":"6% more sellable yield and 17% more net profit for the winning algorithm","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC7698269/","https://doi.org/10.3390/s20226430","https://doi.org/10.3390/s19081807","https://www.greenhousegrower.com/technology/autonomous-greenhouse-challenge-winner-on-the-power-of-artificial-intelligence/"],"source_refs":["AIBD-0200-source-0001","AIBD-0200-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"winning algorithm raised sellable yield 6% and net profit 17% over reference growers","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"winning algorithm raised sellable yield 6% and net profit 17% over reference growers","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0200-reported-measure","evaluation_data":null},{"case_id":"AIBD-0364","case_url":"records/AIBD-0364.html","measure":null,"reported_result":null,"baseline":"a bad crash in a normal car without Autopilot, by the driver's account of an earlier event","result":"the car straightened its course after his swerve, by his account","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.dailystar.co.uk/news/world-news/elon-musk-says-tesla-car-22163770","https://www.dhnet.be/actu/belgique/un-boulanger-belge-fait-reagir-elon-musk-en-personne-suite-a-l-exploit-de-sa-tesla-le-systeme-est-surhumain-5ee2512d9978e21bd06f5947","https://electrek.co/2020/06/08/tesla-autopilot-avoid-pig-video/","https://www.lavenir.net/cnt/dmf20200610_01481837/video-un-belge-evite-un-sanglier-grace-a-sa-tesla-et-fait-reagir-elon-musk-en-personne","https://www.lefigaro.fr/flash-actu/le-conducteur-d-une-tesla-evite-de-justesse-un-sanglier-grace-a-l-autopilot-20200612","https://oecd.ai/en/incidents/2020-06-10-258f","https://www.tvl.be/nieuws/elon-musk-reageert-op-tesla-van-houthalense-bakker-die-uitwijkt-voor-everzwijn-99790"],"source_refs":["AIBD-0364-source-0002","AIBD-0364-source-0003","AIBD-0364-source-0001","AIBD-0364-source-0004","AIBD-0364-source-0005","AIBD-0364-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0364-reported-measure","evaluation_data":null},{"case_id":"AIBD-0629","case_url":"records/AIBD-0629.html","measure":"PPV, NPV, sensitivity, specificity, no-result rate.","reported_result":"a positive predictive value of 81%","baseline":"Reference diagnosis","result":"PPV 81%, NPV 98%, sensitivity 98%, specificity 79% among determinate outputs","baseline_value":null,"result_value":null,"unit":null,"effect":"PPV 81%, NPV 98%, sensitivity 98%, specificity 79%; no result for 68%.","design":"other.","sample":"711 enrolled, 425 completers at 14 US sites.","uncertainty":"No result for 68% of children; high dropout; outputs blinded so care was unchanged.","comparison_basis":"measured","source_urls":["https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN200069.pdf","https://pmc.ncbi.nlm.nih.gov/articles/PMC9072329"],"source_refs":["AIBD-0629-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Specialist reference diagnosis.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"PPV 81%, NPV 98%, sensitivity 98%, specificity 79%; no result for 68%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One assessment per child.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Prospective double-blinded single-arm diagnostic accuracy study against a DSM-5 specialist reference standard.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"PPV, NPV, sensitivity, specificity, no-result rate.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"ClinicalTrials.gov NCT04151290.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"711 enrolled, 425 completers at 14 US sites.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"No result for 68% of children; high dropout; outputs blinded so care was unchanged.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0629-reported-measure","evaluation_data":null},{"case_id":"AIBD-0024","case_url":"records/AIBD-0024.html","measure":"recorded a peak resting pulse rate of 188 beats per minute","reported_result":"recorded a peak resting pulse rate of 188 beats per minute","baseline":"Four years of palpitations without a diagnosis, and an emergency visit on 9 June 2020 that ended with symptomatic treatment","result":"A recorded 188 bpm episode that, with a routine ECG showing WPW the next day, supported referral and catheter ablation on 29 June 2020","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://europepmc.org/article/PMC/PMC8075737","https://europepmc.org/article/PMC/PMC11522799","https://europepmc.org/article/PMC/PMC11083549","https://pmc.ncbi.nlm.nih.gov/articles/PMC13118077/","https://www.accessdata.fda.gov/cdrh_docs/pdf20/K201168.pdf"],"source_refs":["AIBD-0024-source-0001","AIBD-0024-source-0002","AIBD-0024-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"recorded a peak resting pulse rate of 188 beats per minute","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"recorded a peak resting pulse rate of 188 beats per minute","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0024-reported-measure","evaluation_data":null},{"case_id":"AIBD-0194","case_url":"records/AIBD-0194.html","measure":"Mean area under the ROC curve (AUC) and mean predicted suitability at withheld presence sites.","reported_result":"the Mauritania model had the highest mean AUC (0.887)","baseline":"known breeding sites not used in model development","result":"mean predicted suitability of 0.85 at the Mauritanian sites and 0.84 at the Saudi sites","baseline_value":null,"result_value":null,"unit":null,"effect":"Mean AUC of 0.887 for Mauritania, 0.884 for Saudi Arabia and 0.820 for Morocco; mean suitability of 0.85 and 0.84 at withheld Mauritanian and Saudi sites.","design":"other.","sample":"5,406 breeding sites from Morocco (367), Mauritania (2,661) and Saudi Arabia (1,703), 1985 to 2020; 28 Kenyan records from 2020 and 647 Sudanese records from 2013 to 2019 used to check projections.","uncertainty":"Presence-only data with no absence records; no field survey; the authors give no hit rate for the Kenyan and Sudanese check records.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC7371632/","https://doi.org/10.1038/s41598-020-68895-2"],"source_refs":["https://pmc.ncbi.nlm.nih.gov/articles/PMC7371632/"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Presence records withheld from model building.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean AUC of 0.887 for Mauritania, 0.884 for Saudi Arabia and 0.820 for Morocco; mean suitability of 0.85 and 0.84 at withheld Mauritanian and Saudi sites.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Cross-validation with a 70/30 split on presence records, plus projection of the Morocco model to presence records in Mauritania and Saudi Arabia that were not used to build it.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean area under the ROC curve (AUC) and mean predicted suitability at withheld presence sites.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"5,406 breeding sites from Morocco (367), Mauritania (2,661) and Saudi Arabia (1,703), 1985 to 2020; 28 Kenyan records from 2020 and 647 Sudanese records from 2013 to 2019 used to check projections.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Presence-only data with no absence records; no field survey; the authors give no hit rate for the Kenyan and Sudanese check records.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0194-reported-measure","evaluation_data":null},{"case_id":"AIBD-0334","case_url":"records/AIBD-0334.html","measure":null,"reported_result":null,"baseline":"una persona tarda al menos un día para leer 30 expedientes","result":"el sistema puede procesar las sentencias en medio digital en menos de 2 minutos","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.ambitojuridico.com/noticias/tecnologia/constitucional-y-derechos-humanos/pretoria-la-nueva-herramienta-tecnologica-que","https://www.dejusticia.org/wp-content/uploads/2021/04/CPC_informe_Colombia.pdf","https://www.semana.com/nacion/articulo/corte-constitucional-implementara-revision-electronica-de-tutelas/683872/","https://web.archive.org/web/20210419173924/https://www.dejusticia.org/conoce-nuestra-investigacion-sobre-pretoria-la-tecnologia-que-incorpora-la-inteligencia-artificial-a-la-corte-constitucional/","https://www.eltiempo.com/justicia/cortes/se-parte-en-dos-la-historia-de-la-tutela-presidenta-de-la-corte-constitucional-455344","https://www.corteconstitucional.gov.co/relatoria/2024/T-323-24.htm","https://www.corteconstitucional.gov.co/Transparencia/publicaciones/Informe_de_Gesti%C3%B3n_%202022_VF.pdf","https://bibliotecavirtual.unl.edu.ar/publicaciones/index.php/Redoeda/article/download/13824/19280"],"source_refs":["AIBD-0334-source-0002","AIBD-0334-source-0003","AIBD-0334-source-0004","AIBD-0334-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: a person reading tutela files by hand to find the chamber's selection criterion","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: una persona tarda al menos un día para leer 30 expedientes. Result: el sistema puede procesar las sentencias en medio digital en menos de 2 minutos.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0334-reported-measure","evaluation_data":null},{"case_id":"AIBD-0294","case_url":"records/AIBD-0294.html","measure":"sensitivity and specificity of a high signal within two days of the onset of an exacerbation.","reported_result":"EOV signals preceded four of five analyzed cases (sensitivity 80%, specificity 90%)","baseline":"the sensitivity and specificity of high RHR signals were 50% and 80%, respectively","result":"4 out of 5 cases (sensitivity=80%, specificity=90%)","baseline_value":null,"result_value":null,"unit":null,"effect":"EOV sensitivity 80% and specificity 90%, with a positive predictive value of 9.3%; the abstract gives 50% and 80% for resting heart rate, and the results section gives 50% and 88%.","design":"other.","sample":"23 patients with PCR-confirmed COVID-19 at JCHO Kanazawa Hospital, August to November 2020; five exacerbation cases analyzed.","uncertainty":"Five exacerbation cases; short baseline data, with each patient's baseline set to the minimum value over the whole study period; device functions not approved as a medical device at the time.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8448084/"],"source_refs":["AIBD-0294-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated. The EOV metric comes from the device, and the signal rule and the decision to order polysomnography came from the investigators.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"high resting-heart-rate signals from the same devices.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"EOV sensitivity 80% and specificity 90%, with a positive predictive value of 9.3%; the abstract gives 50% and 80% for resting heart rate, and the results section gives 50% and 88%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Fitbit Charge 3 worn for up to 30 days (median 19 days).","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective comparison of investigator-defined high-EOV and high resting-heart-rate signals against recorded exacerbation events.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"sensitivity and specificity of a high signal within two days of the onset of an exacerbation.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"23 patients with PCR-confirmed COVID-19 at JCHO Kanazawa Hospital, August to November 2020; five exacerbation cases analyzed.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Five exacerbation cases; short baseline data, with each patient's baseline set to the minimum value over the whole study period; device functions not approved as a medical device at the time.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0294-reported-measure","evaluation_data":null},{"case_id":"AIBD-0066","case_url":"records/AIBD-0066.html","measure":"Adjusted absolute difference in in-hospital mortality.","reported_result":"3.3% absolute (18.7% relative) lower in-hospital mortality","baseline":"patients whose alert was not confirmed by a provider within 3 h","result":"reduced in-hospital mortality rate (3.3%, confidence interval (CI) 1.7, 5.1%, adjusted absolute reduction, and 18.7%, CI 9.4, 27.0%, adjusted relative reduction)","baseline_value":null,"result_value":null,"unit":null,"effect":"Unadjusted mortality 14.6% against 19.2%; adjusted difference -3.3% (CI -5.1 to -1.7).","design":"controlled observational.","sample":"Five Johns Hopkins hospitals, 2018 to September 2020; 590,736 encounters monitored; 6,877 sepsis encounters analyzed (4,220 confirmed within 3 hours, 2,657 not).","uncertainty":"Observational comparison within alerted patients, with no adjustment for provider-level features. A reanalysis by University of California San Diego researchers argues that confounding by processes of care biases the result.","comparison_basis":"measured","source_urls":["https://europepmc.org/abstract/MED/35864252","https://api.biorxiv.org/details/medrxiv/10.1101/2022.08.08.22278526","https://pmc.ncbi.nlm.nih.gov/articles/PMC9512693/","https://www.accessdata.fda.gov/cdrh_docs/pdf25/K250680.pdf"],"source_refs":["AIBD-0066-source-0004"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Alerted sepsis patients whose alert was not confirmed within 3 hours.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Unadjusted mortality 14.6% against 19.2%; adjusted difference -3.3% (CI -5.1 to -1.7).","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Adjusted in-hospital mortality, change in organ failure score and length of stay, by whether a provider confirmed the alert within 3 hours.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Adjusted absolute difference in in-hospital mortality.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Five Johns Hopkins hospitals, 2018 to September 2020; 590,736 encounters monitored; 6,877 sepsis encounters analyzed (4,220 confirmed within 3 hours, 2,657 not).","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Observational comparison within alerted patients, with no adjustment for provider-level features. A reanalysis by University of California San Diego researchers argues that confounding by processes of care biases the result.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0066-reported-measure","evaluation_data":null},{"case_id":"AIBD-0323","case_url":"records/AIBD-0323.html","measure":"an 85 percent cooperation rate overall by fleets","reported_result":"an 85 percent cooperation rate overall by fleets","baseline":"the baseline of transits in the area in 2016","result":"40% less risk to whales during the 2025 VSR season","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://bluewhalesblueskies.org/impact/reducing-ship-strikes/","https://news.mongabay.com/2020/09/new-tool-alerts-ships-when-whales-are-near-but-will-they-slow-down/","https://news.ucsb.edu/2022/020717/safe-harbor","https://www.kcbx.org/environment-and-energy/2024-04-11/ucsbs-whale-collision-prevention-tool-reveals-speed-limit-violations-by-large-ships","https://www.independent.com/2026/07/16/record-60-shipping-lines-take-part-in-slowdown-program/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13419462/","https://whalesafe.com/methodology/"],"source_refs":["AIBD-0323-source-0001","AIBD-0323-source-0002","AIBD-0323-source-0003","AIBD-0323-source-0004","AIBD-0323-source-0005","AIBD-0323-source-0006","AIBD-0323-source-0007"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the 2016 baseline of vessel transits in the area, under the BWBS voluntary speed program that Whale Safe supports (not Whale Safe alone)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: the baseline of transits in the area in 2016. Result: 40% less risk to whales during the 2025 VSR season.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"an 85 percent cooperation rate overall by fleets","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0323-reported-measure","evaluation_data":null},{"case_id":"AIBD-0032","case_url":"records/AIBD-0032.html","measure":null,"reported_result":null,"baseline":"Missing from Allahabad since 14 July 2015 and in a child welfare center at Goalpara since 23 July 2015, unidentified","result":"Matched by Darpan and recognized in person by his parents by 9 October 2020","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.tribuneindia.com/news/nation/face-recognition-tool-reunites-up-boy-with-parents-after-5-years-153319","https://www.etvbharat.com/english/bharat/bharat-news/5-years-on-missing-up-boy-traced-to-assam-through-darpan/na20201009182148536","https://telanganatoday.com/ts-police-darpan-helps-child-from-up-reunite-with-parents-after-5-years","https://www.thehansindia.com/telangana/telangana-polices-darpan-helps-mp-boy-to-reunite-with-parents-after-10-years-661805","https://www.aninews.in/news/national/general-news/darpan-face-recognition-portal-helps-reunite-33-missing-children-with-parents20201012233712/"],"source_refs":["AIBD-0032-source-0001","AIBD-0032-source-0002","AIBD-0032-source-0003","AIBD-0032-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0032-reported-measure","evaluation_data":null},{"case_id":"AIBD-0664","case_url":"records/AIBD-0664.html","measure":"Resolutions per hour.","reported_result":"raised issues resolved per hour by 15% on average","baseline":"Issues resolved per hour without access","result":"15% more issues resolved per hour on average","baseline_value":null,"result_value":null,"unit":null,"effect":"14% on average and 34% for novice and low-skilled agents (working paper); 15% on average (published).","design":"quasi-experimental.","sample":"5,179 agents and about 3 million chats in the working paper; 5,172 agents in the published version.","uncertainty":"One firm; staggered rollout chosen by the firm; attrition estimates lack agent fixed effects; small quality declines for the most skilled agents.","comparison_basis":"measured","source_urls":["https://ideas.repec.org/a/oup/qjecon/v140y2025i2p889-942..html","https://www.nber.org/system/files/working_papers/w31161/w31161.pdf"],"source_refs":["AIBD-0664-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"About 14% to 15% more resolutions per hour.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Agents not yet given access, with agent and time fixed effects.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"14% on average and 34% for novice and low-skilled agents (working paper); 15% on average (published).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Access to real-time AI suggestions during chats.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Issues resolved per hour, resolution rate, handle time, net promoter score and attrition from the firm's systems.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Resolutions per hour.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Firm data not public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"5,179 agents and about 3 million chats in the working paper; 5,172 agents in the published version.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One firm; staggered rollout chosen by the firm; attrition estimates lack agent fixed effects; small quality declines for the most skilled agents.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0664-reported-measure","evaluation_data":null},{"case_id":"AIBD-0093","case_url":"records/AIBD-0093.html","measure":"Food security, mental health and self-perceived economic status.","reported_result":"reduced exclusion errors by 4–21% relative to geographic alternatives","baseline":"errors of exclusion of 59%–78% for the feasible alternatives","result":"errors of exclusion of 53% for the phone-based approach, a reduction of 4–21%","baseline_value":null,"result_value":null,"unit":null,"effect":"Survey data show gains of 0.06 standard deviations in food security (p = 0.003), 0.07 in mental health (p < 0.001) and 0.04 in self-perceived socioeconomic status (p = 0.074); phone-data estimates did not show similar effects.","design":"randomized.","sample":"Eligible subscribers in the first phase of the GiveDirectly Novissi program, which covered Togo's 100 poorest cantons, with treatment randomly assigned at the individual level.","uncertainty":"Written by the team that built the targeting; effects on financial health, financial inclusion and health care access were positive but not statistically significant; phone-data impact estimates disagree with survey estimates.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8967719/","https://documents.worldbank.org/curated/en/099751009222330502/pdf/IDU0e83f857301ff1047bf082710a8d21ddf42c3.pdf","https://www.givedirectly.org/togo/","https://www.npr.org/sections/goatsandsoda/2021/02/15/966848542/the-pandemic-pushed-this-farmer-into-deep-poverty-then-something-amazing-happene","https://www.nber.org/papers/w31751"],"source_refs":["AIBD-0093-source-0005"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not measured: the trial tests the transfers and does not compare the machine-learning targeting with another method.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Eligible subscribers randomly assigned to a control group whose payments began in June 2021, after the endline survey.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Survey data show gains of 0.06 standard deviations in food security (p = 0.003), 0.07 in mental health (p < 0.001) and 0.04 in self-perceived socioeconomic status (p = 0.074); phone-data estimates did not show similar effects.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mobile-money transfers paid to registrants chosen by satellite and phone-data targeting.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Endline phone survey of treatment and control subscribers in May 2021 by enumerators of Togo's national statistics institute (INSEED), compared with welfare predicted from mobile phone data.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Food security, mental health and self-perceived economic status.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Eligible subscribers in the first phase of the GiveDirectly Novissi program, which covered Togo's 100 poorest cantons, with treatment randomly assigned at the individual level.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Written by the team that built the targeting; effects on financial health, financial inclusion and health care access were positive but not statistically significant; phone-data impact estimates disagree with survey estimates.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0093-reported-measure","evaluation_data":null},{"case_id":"AIBD-0333","case_url":"records/AIBD-0333.html","measure":"more than 10 million pages processed by May 2021","reported_result":"more than 10 million pages processed by May 2021","baseline":"três horas","result":"cinco segundos","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://noticias.stf.jus.br/postsnoticias/ministra-carmen-lucia-anuncia-inicio-de-funcionamento-do-projeto-victor-de-inteligencia-artificial/","https://noticias.stf.jus.br/postsnoticias/projeto-victor-avanca-em-pesquisa-e-desenvolvimento-para-identificacao-dos-temas-de-repercussao-geral/","https://www.cnj.jus.br/inteligencia-artificial-trabalho-judicial-de-40-minutos-pode-ser-feito-em-5-segundos/","https://noticias.stf.jus.br/postsnoticias/stf-finaliza-testes-de-nova-ferramenta-de-inteligencia-artificial/","https://web.archive.org/web/2026/https://www.e-publicacoes.uerj.br/index.php/quaestioiuris/article/view/52714","https://revistas.ufpr.br/rinc/article/view/e252"],"source_refs":["AIBD-0333-source-0001","AIBD-0333-source-0002","AIBD-0333-source-0003","AIBD-0333-source-0004","AIBD-0333-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: a court clerk converting image files to text by hand","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: três horas. Result: cinco segundos.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"more than 10 million pages processed by May 2021","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0333-reported-measure","evaluation_data":null},{"case_id":"AIBD-0221","case_url":"records/AIBD-0221.html","measure":"Accuracy, recall and precision for the dyslexia class.","reported_result":"79.8% accuracy with 80.6% recall and 79.3% precision","baseline":"89.2% accuracy with no child with dyslexia detected","result":"79.8% accuracy with 80.6% recall and 79.3% precision for the dyslexia class","baseline_value":null,"result_value":null,"unit":null,"effect":"79.8% accuracy, 80.6% recall and 79.3% precision; 78.1% recall on tablets for ages 12 to 17.","design":"other.","sample":"3,644 Spanish-speaking participants aged 7 to 17 (392 with a diagnosis) recruited through 113 organizations in Argentina, Chile, Colombia, Spain and the United States; tablet sample of 1,395.","uncertainty":"Single dataset with no external validation; the makers wrote the study and hold a patent application. Independent re-analyses in 2026 report a demographic confound by native language.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC7710040/","https://europepmc.org/article/PMC/PMC13656794","https://pmc.ncbi.nlm.nih.gov/articles/PMC13580141/"],"source_refs":["AIBD-0221-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"A trivial classifier that labels everyone as not having dyslexia (89.2% accuracy, no detection).","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"79.8% accuracy, 80.6% recall and 79.3% precision; 78.1% recall on tablets for ages 12 to 17.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Diagnostic accuracy of a random-forest classifier on 32 gamified exercises, scored against professional dyslexia diagnoses with 10-fold cross-validation, and a second test on tablets.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Accuracy, recall and precision for the dyslexia class.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Dataset public on Kaggle (doi 10.34740/kaggle/dsv/1617514).","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"3,644 Spanish-speaking participants aged 7 to 17 (392 with a diagnosis) recruited through 113 organizations in Argentina, Chile, Colombia, Spain and the United States; tablet sample of 1,395.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Single dataset with no external validation; the makers wrote the study and hold a patent application. Independent re-analyses in 2026 report a demographic confound by native language.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0221-reported-measure","evaluation_data":null},{"case_id":"AIBD-0001","case_url":"records/AIBD-0001.html","measure":"Walking distance and speed, balance, assistance required, muscle activity, and decoder accuracy.","reported_result":"walk more than 10 meters with crutches while the stimulation was off","baseline":"after three years of regular training with stimulation only, he had reached a plateau of recovery; EES only condition showed significantly shorter step length due to increase of failed steps; WISCI II score 6 before STIMO","result":"the BSI enabled walking with gait features that were markedly closer to those quantified in healthy individuals; WISCI II 16 after STIMO-BSI","baseline_value":null,"result_value":null,"unit":null,"effect":"After rehabilitation, the participant walked more than 10 meters with crutches and no stimulation. Brain-controlled stimulation also improved walking across uneven terrain.","design":"Researchers followed one participant and compared walking under different stimulation conditions. Six independent physiotherapists assessed gait videos without knowing the condition.","sample":"One 38-year-old man at CHUV in Lausanne, ten years after an incomplete C5/C6 spinal cord injury.","uncertainty":"The study involved one participant. Surgery, stimulation, and rehabilitation accompanied the decoder, so their individual contributions to recovery were not isolated.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10232367/","https://actu.epfl.ch/news/thought-controlled-walking-again-after-spinal-co-4","https://www.cbsnews.com/news/paralyzed-man-walks-again-using-implants-connecting-brain-with-spinal-cord/","https://www.globenewswire.com/news-release/2024/09/19/2948718/0/en/ONWARD-Medical-Announces-Third-Implant-of-Brain-Computer-Interface-BCI-System-to-Restore-Movement-after-Spinal-Cord-Injury.html","https://ir.onwd.com/static-files/48e1c5b9-e843-4895-ac81-e0e936a841e2","https://zenodo.org/records/7680471"],"source_refs":["AIBD-0001-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Brain-controlled stimulation produced more natural steps than stimulation triggered by motion sensors. The study did not compare different decoders.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Walking with brain-controlled stimulation was compared with stimulation triggered by motion sensors and with no stimulation.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Researchers followed one participant and compared walking under different stimulation conditions. Six independent physiotherapists assessed gait videos without knowing the condition.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"After rehabilitation, the participant walked more than 10 meters with crutches and no stimulation. Brain-controlled stimulation also improved walking across uneven terrain.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One participant completed 40 rehabilitation sessions over 15 weeks, followed by seven months of home use.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"STIMO-BSI, NCT04632290.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Researchers recorded movements and muscle activity, performed clinical walking tests, and collected the participant’s reports.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Walking distance and speed, balance, assistance required, muscle activity, and decoder accuracy.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The authors deposited source data and supporting MATLAB scripts in Zenodo.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One 38-year-old man at CHUV in Lausanne, ten years after an incomplete C5/C6 spinal cord injury.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The study involved one participant. Surgery, stimulation, and rehabilitation accompanied the decoder, so their individual contributions to recovery were not isolated.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0001-reported-measure","evaluation_data":null},{"case_id":"AIBD-0083","case_url":"records/AIBD-0083.html","measure":"captured R$426m in 2021 (R$194m in 2020)","reported_result":"captured R$426m in 2021 (R$194m in 2020)","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://portal.tcu.gov.br/imprensa/noticias/atuacao-do-tcu-gera-recorde-de-beneficios-financeiros-em-2021","https://portal.tcu.gov.br/imprensa/noticias/tcu-ministerio-da-infraestrutura-e-dnit-firmam-acordo-para-aperfeicoar-fiscalizacao-e-prevenir-fraudes-em-licitacoes","https://portal.tcu.gov.br/imprensa/noticias/uso-de-inteligencia-artificial-aprimora-processos-internos-no-tcu.htm","https://inforchannel.com.br/2017/09/27/tcu-cria-robo-para-varredura-diaria-de-irregularidades-em-editais/","https://www.mpc.es.gov.br/2018/03/como-as-robos-alice-sofia-e-monica-ajudam-o-tcu-a-cacar-irregularidades-em-licitacoes/","https://zenite.blog.br/tcu-aponta-irregularidades-em-contratacoes-publicas-e-destaca-avanco-do-alice-360-com-ia/","https://pmc.ncbi.nlm.nih.gov/articles/PMC10039446/"],"source_refs":["AIBD-0083-source-0001","AIBD-0083-source-0002","AIBD-0083-source-0003","AIBD-0083-source-0004","AIBD-0083-source-0005","AIBD-0083-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"captured R$426m in 2021 (R$194m in 2020)","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"captured R$426m in 2021 (R$194m in 2020)","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0083-reported-measure","evaluation_data":null},{"case_id":"AIBD-0094","case_url":"records/AIBD-0094.html","measure":"49,000 beneficiary households as of 15 October 2021","reported_result":"49,000 beneficiary households as of 15 October 2021","baseline":"0.79 precision at the ward level (survey benchmark)","result":"0.92 precision at the ward level (ML-based maps)","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC9371736/","https://documents.worldbank.org/curated/en/717321706545772846/pdf/Nigeria-National-Social-Safety-Net-Program-Scale-Up-Project.pdf","https://documents.worldbank.org/curated/en/099012925120031079/pdf/P17693518b066e0ff1b62c194b8b78c78dd.pdf","https://www.hrw.org/report/2021/07/28/between-hunger-and-the-virus/the-impact-of-the-covid-19-pandemic-on-people-living","https://blogs.worldbank.org/en/opendata/using-big-data-and-machine-learning-locate-poor-nigeria"],"source_refs":["AIBD-0094-source-0002","AIBD-0094-source-0003","AIBD-0094-source-0001","AIBD-0094-source-0004","AIBD-0094-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"survey-based poverty map benchmark (DHS/NLSS data)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 0.79 precision at the ward level (survey benchmark). Result: 0.92 precision at the ward level (ML-based maps).","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"49,000 beneficiary households of the government's COVID-CT as of 15 October 2021","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0094-reported-measure","evaluation_data":null},{"case_id":"AIBD-0121","case_url":"records/AIBD-0121.html","measure":"Phantom limb pain and stump pain intensity on a 0 to 10 scale; EQ-5D-5L index.","reported_result":"her phantom limb pain fell from 5 to 3","baseline":"phantom pain level 5","result":"phantom pain level 3; stump pain vanished entirely","baseline_value":null,"result_value":null,"unit":null,"effect":"Phantom limb pain fell from 5 to 3; stump pain fell from 6 to 0; EQ-5D-5L rose from 0.23 to 0.63.","design":"case study.","sample":"One woman with a right transradial amputation, treated at Sahlgrenska University Hospital, Sweden; study from September 2018 to April 2021.","uncertainty":"One patient, no control group, no blinding; the authors include developers and shareholders of Integrum AB and Prensilia SRL.","comparison_basis":"measured","source_urls":["https://keyt.com/health/cnn-health/2023/10/11/its-not-star-wars-level-tech-yet-but-doctors-get-a-step-closer-to-a-bionic-hand-with-special-surgery-and-ai","https://doi.org/10.1126/scirobotics.adf7360","https://medicalxpress.com/news/2023-10-bionic-merges-user-nervous-skeletal.html","https://www.popsci.com/technology/bionic-hand-phantom-pain","https://gizmodo.com/bionic-mia-hand-prosthetic-amputee-phantom-pain-1850919751","https://newatlas.com/medical/bionic-hand-restores-life-amputees","https://www.sciencealert.com/revolutionary-bionic-hand-fuses-with-womans-bones-muscles-and-nerves","https://www.detop-project.eu/2019/02/05/the-first-dexterous-and-sentient-hand-prosthesis-has-been-successfully-implanted/"],"source_refs":["https://research.chalmers.se/publication/537965/file/537965_Fulltext.pdf"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated. The paper attributes the outcome to the whole implant, surgery and electrode system and does not test a learned decoder against an alternative.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The same patient 6 weeks before the intervention, when she used a socket prosthesis with surface electrodes.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Phantom limb pain fell from 5 to 3; stump pain fell from 6 to 0; EQ-5D-5L rose from 0.23 to 0.63.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Assessments 6 weeks before and 123 weeks after the intervention; by publication she had used the hand in daily life for more than 3 years.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Patient-reported questionnaires (Questionnaire for Phantom Limb Pain Tracking, EQ-5D-5L, DASH, Q-ULA) and function tests (ACMC, SHAP) given before and after the intervention.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Phantom limb pain and stump pain intensity on a 0 to 10 scale; EQ-5D-5L index.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One woman with a right transradial amputation, treated at Sahlgrenska University Hospital, Sweden; study from September 2018 to April 2021.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One patient, no control group, no blinding; the authors include developers and shareholders of Integrum AB and Prensilia SRL.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0121-reported-measure","evaluation_data":null},{"case_id":"AIBD-0638","case_url":"records/AIBD-0638.html","measure":"Discrepancy rate; length of stay.","reported_result":"the rate of discrepancies between emergency physicians' and radiologists' readings fell by 17%","baseline":"Discrepancy rate and length of stay in January to June 2020","result":"Discrepancy rate 17% lower and mean stay 9 minutes shorter in January to June 2021","baseline_value":null,"result_value":null,"unit":null,"effect":"Discrepancy rate down 17%; mean stay down 9 minutes.","design":"before/after.","sample":"3,720 adult trauma patients with limb radiographs (1,780 in 2020, 1,940 in 2021) at one level I trauma center.","uncertainty":"Retrospective, single center, no concurrent control; clinically relevant discrepancies did not change significantly.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11433213/"],"source_refs":["AIBD-0638-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"17% relative fall in discrepancies.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Same six months of the previous year.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Discrepancy rate down 17%; mean stay down 9 minutes.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"All eligible patients in the 2021 period.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective comparison of emergency physician and radiologist readings, length of stay and consultations from records.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Discrepancy rate; length of stay.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not stated.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"3,720 adult trauma patients with limb radiographs (1,780 in 2020, 1,940 in 2021) at one level I trauma center.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective, single center, no concurrent control; clinically relevant discrepancies did not change significantly.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0638-reported-measure","evaluation_data":null},{"case_id":"AIBD-0324","case_url":"records/AIBD-0324.html","measure":"995 door/window alerts from January to May 2021","reported_result":"995 door/window alerts from January to May 2021","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://oecd.ai/en/incidents/2021-02-25-adca","https://www.sohu.com/a/452580009_120823584","https://www.163.com/dy/article/G3NCP8NI0514EGPO.html","https://tech.gmw.cn/2021-05/07/content_34826280.htm","https://m.thepaper.cn/newsDetail_forward_13463904","https://www.cac.gov.cn/2021-03/04/c_1616435886725223.htm","https://www.smartcity.team/cases/smartcitycases/shanghaichengshizuixiaoguanlidanyuan/","http://big-data.zhiding.cn/2021/0225/3132347.shtml"],"source_refs":["AIBD-0324-source-0001","AIBD-0324-source-0002","AIBD-0324-source-0003","AIBD-0324-source-0004","AIBD-0324-source-0005","AIBD-0324-source-0006","AIBD-0324-source-0007","AIBD-0324-source-0008"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"995 door/window alerts from January to May 2021","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"995 door/window alerts from January to May 2021","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0324-reported-measure","evaluation_data":null},{"case_id":"AIBD-0042","case_url":"records/AIBD-0042.html","measure":"The primary outcomes were recovery time in ACTT-2, disease progression in COV-BARRIER, and deaths within 28 days in RECOVERY. The economists estimated costs and quality-adjusted life-years.","reported_result":"mortality was 12% with the drug and 14% with usual care alone","baseline":"8 days (95% CI, 7 to 9)","result":"7 days (95% confidence interval [CI], 6 to 8)","baseline_value":null,"result_value":null,"unit":null,"effect":"Median recovery was 7 rather than 8 days in ACTT-2. Deaths by day 28 were 8% versus 13% in COV-BARRIER and 12% versus 14% in RECOVERY.","design":"Investigators randomly assigned treatment in all three trials. ACTT-2 and COV-BARRIER concealed assignments, while RECOVERY disclosed them. The economic study simulated costs and health.","sample":"The trials enrolled hospitalized patients in 2020–2021: 1,033 in ACTT-2, 1,525 in COV-BARRIER, and 8,156 in RECOVERY. The economic study modeled insured U.S. adults.","uncertainty":"COV-BARRIER investigators did not establish less disease progression, which prevented confirmatory testing of mortality under their analysis plan. RECOVERY participants knew their treatment. Lilly funded COV-BARRIER and the economic study. None compared discovery with and without AI.","comparison_basis":"measured","source_urls":["https://europepmc.org/article/PMC/PMC7745180","https://pmc.ncbi.nlm.nih.gov/articles/PMC8409066/","https://pmc.ncbi.nlm.nih.gov/articles/PMC9333998/","https://europepmc.org/article/PMC/PMC7137985","https://www.frontiersin.org/journals/pharmacology/articles/10.3389/fphar.2021.709856/full","https://www.fda.gov/news-events/press-announcements/coronavirus-covid-19-update-fda-authorizes-drug-combination-treatment-covid-19","https://www.who.int/news/item/14-01-2022-who-recommends-two-new-drugs-to-treat-covid-19","https://www.benevolent.com/news-and-media/press-releases-and-in-media/benevolentai-unaudited-preliminary-results-year-ended-31-december-2022/","https://www.benevolent.com/news-and-media/press-releases-and-in-media/benevolentais-platform-derived-hypothesis-covid-19-treatment-validated-us-niaid-randomised-control-trial/","https://pmc.ncbi.nlm.nih.gov/articles/PMC8487786/","https://www.fda.gov/media/144473/download","https://www.accessdata.fda.gov/drugsatfda_docs/appletter/2022/207924Orig1s006ltr.pdf","https://www.ema.europa.eu/en/medicines/human/withdrawn-applications/olumiant"],"source_refs":["AIBD-0042-source-0005","AIBD-0042-source-0001","AIBD-0042-source-0002","AIBD-0042-source-0010"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The software probably reduced the work needed to assemble and examine relevant biomedical findings.","editor_note":"Smith and colleagues report about 40,000 added relationships and a starting network of 556 proteins (reference 4). These quantities support an inference of substantial assistance, but the authors did not time a comparable search without AI.","inferred":true}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Patients received baricitinib or no baricitinib alongside the same background care. ACTT-2 gave remdesivir to both groups.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Investigators randomly assigned treatment in all three trials. ACTT-2 and COV-BARRIER concealed assignments, while RECOVERY disclosed them. The economic study simulated costs and health.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Median recovery was 7 rather than 8 days in ACTT-2. Deaths by day 28 were 8% versus 13% in COV-BARRIER and 12% versus 14% in RECOVERY.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"ACTT-2 assigned 515 patients to baricitinib and 518 to control. COV-BARRIER assigned 764 and 761. RECOVERY assigned 4,148 and 4,008.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"ACTT-2, COV-BARRIER, RECOVERY, and Ohsfeldt and colleagues’ 2021 economic study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Investigators used clinical assessments, follow-up forms, and health records. The economists combined trial findings with treatment prices and costs of care.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"The primary outcomes were recovery time in ACTT-2, disease progression in COV-BARRIER, and deaths within 28 days in RECOVERY. The economists estimated costs and quality-adjusted life-years.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Trial registrations, protocols, analysis plans, and appendices are public. The complete historical AI system and economic model are unavailable, and some cost inputs are unpublished.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"The trials enrolled hospitalized patients in 2020–2021: 1,033 in ACTT-2, 1,525 in COV-BARRIER, and 8,156 in RECOVERY. The economic study modeled insured U.S. adults.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"COV-BARRIER investigators did not establish less disease progression, which prevented confirmatory testing of mortality under their analysis plan. RECOVERY participants knew their treatment. Lilly funded COV-BARRIER and the economic study. None compared discovery with and without AI.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0042-reported-measure","evaluation_data":{"ACTT-2":{"metric":"Time to recovery","role":"Primary outcome","period_days":28,"baricitinib_median_days":7,"control_median_days":8},"COV-BARRIER":{"primary":"Progression to high-flow oxygen, noninvasive ventilation, invasive ventilation, ECMO, or death by day 28.","primary_values_percent":[27.8,30.5],"mortality_role":"Prespecified key secondary outcome","mortality_days":28,"baricitinib_deaths":62,"baricitinib_randomized":764,"control_deaths":100,"control_randomized":761},"RECOVERY":{"metric":"All-cause mortality","role":"Primary outcome","period_days":28,"baricitinib_deaths":514,"baricitinib_randomized":4148,"control_deaths":546,"control_randomized":4008},"economic_model":{"metrics":["Incremental hospital expenditure","Incremental reimbursement","Incremental net hospital costs","Incremental lifetime payer costs","Quality-adjusted life-years","Cost per quality-adjusted life-year"]}}},{"case_id":"AIBD-0028","case_url":"records/AIBD-0028.html","measure":"Time in range, time below range and heavy-for-date infants.","reported_result":"hypoglycemic episodes fell from 2.1–2.5 to 1.4 per day","baseline":"Hypoglycemic episodes (/day) 2.5 and 2.1; time with AUC below 70 mg/dL 1.1% and 1.4%","result":"Hypoglycemic episodes (/day) 1.4; mean AUC < 70 mg/dL of 0.5%","baseline_value":null,"result_value":null,"unit":null,"effect":"One heavy-for-date infant among 11 on the pump against five among 11 without it; time in range above 70% in the second and third trimesters on the pump.","design":"controlled observational.","sample":"22 women with type 1 diabetes who delivered between 2011 and 2021 at two Tokyo hospitals; 11 used the MiniMed 640G with predictive low glucose suspend.","uncertainty":"Small retrospective sample with historical controls who lacked real-time continuous glucose monitoring.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8172753/","https://doi.org/10.1007/s13340-020-00489-9","https://europepmc.org/article/PMC/PMC11886589","https://pmc.ncbi.nlm.nih.gov/articles/PMC11291783/"],"source_refs":["AIBD-0028-source-0004"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Longer insulin suspension went with higher time in range (r = 0.62) and lower time below range (r = -0.40).","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"11 pregnancies managed without a sensor-augmented pump before its 2018 launch in Japan: ten on multiple daily injections and one on a pump without a sensor.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"One heavy-for-date infant among 11 on the pump against five among 11 without it; time in range above 70% in the second and third trimesters on the pump.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective review of glucose records, insulin suspension time and neonatal outcomes.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time in range, time below range and heavy-for-date infants.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"22 women with type 1 diabetes who delivered between 2011 and 2021 at two Tokyo hospitals; 11 used the MiniMed 640G with predictive low glucose suspend.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small retrospective sample with historical controls who lacked real-time continuous glucose monitoring.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0028-reported-measure","evaluation_data":null},{"case_id":"AIBD-0051","case_url":"records/AIBD-0051.html","measure":"T5 typed 196 correct characters in 14.6 minutes (13.4 per minute)","reported_result":"T5 typed 196 correct characters in 14.6 minutes (13.4 per minute)","baseline":null,"result":"Communication bitrates were equivalent between cabled and wireless configurations.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8218873/","https://europepmc.org/abstract/MED/30462658"],"source_refs":["AIBD-0051-source-0001","AIBD-0051-source-0002","AIBD-0051-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the cabled intracortical BCI","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Result: Communication bitrates were equivalent between cabled and wireless configurations.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"T5 typed 196 correct characters in 14.6 minutes (13.4 per minute)","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0051-reported-measure","evaluation_data":null},{"case_id":"AIBD-0103","case_url":"records/AIBD-0103.html","measure":null,"reported_result":null,"baseline":"collision counts with the full lighted window area","result":"about 11 times fewer collisions in spring and 6 times fewer in fall when lighted window area is halved","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8214683/","https://birdshouston.org/conservation/lights-out.html","https://birdcast.org/migration-tools/migration-forecast-maps/","https://web.archive.org/web/2026/https://www.houstonchronicle.com/lifestyle/home-garden/article/A-darker-Houston-sky-is-the-goal-of-Lights-Out-16106554.php"],"source_refs":["AIBD-0103-source-0001","AIBD-0103-source-0002","AIBD-0103-source-0003","AIBD-0103-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0103-reported-measure","evaluation_data":null},{"case_id":"AIBD-0071","case_url":"records/AIBD-0071.html","measure":"two statistically distinct groups, with a transition in columns 27–29","reported_result":"two statistically distinct groups, with a transition in columns 27–29","baseline":"the entire manuscript was copied by one scribe, with minor interventions by other, contemporaneous and also much later, scribes","result":"two main scribes, each showing different writing patterns, were responsible for the Great Isaiah Scroll","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0249769","https://phys.org/news/2021-04-code-dead-sea-scrolls.html","https://theconversation.com/dead-sea-scrolls-two-scribes-probably-wrote-one-of-the-manuscripts-159169","https://library.imaging.org/archiving/articles/22/1/14"],"source_refs":["AIBD-0071-source-0001","AIBD-0071-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the prevailing scholarly view that a single scribe copied the whole scroll","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: the entire manuscript was copied by one scribe, with minor interventions by other, contemporaneous and also much later, scribes. Result: two main scribes, each showing different writing patterns, were responsible for the Great Isaiah Scroll.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"two scribes wrote the Great Isaiah Scroll, switching at columns 27–29","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0071-reported-measure","evaluation_data":null},{"case_id":"AIBD-0099","case_url":"records/AIBD-0099.html","measure":"Cumulative weekly drops in listening engagement.","reported_result":"reduced cumulative engagement drops by 32.0% compared with messaging alone","baseline":"Cumulative weekly engagement drops in the comparison groups","result":"32.0% fewer versus messaging alone (p = .044); 28.3% fewer versus rotation (p = .098)","baseline_value":null,"result_value":null,"unit":null,"effect":"Engagement drops fell by 32.0% versus regular messages alone and 28.3% versus round-robin calling.","design":"Researchers randomly assigned participants to model-selected calls, round-robin calls, or regular messaging alone.","sample":"The groups contained 7,668, 7,668, and 7,667 participants, respectively.","uncertainty":"The study measured listening, not clinical health. Cumulative drops can count the same participant in more than one week.","comparison_basis":"measured","source_urls":["https://ojs.aaai.org/index.php/AAAI/article/view/21460","https://research.google/blog/using-ml-to-boost-engagement-with-a-maternal-and-child-health-program-in-india/","https://arxiv.org/abs/2604.07384","https://arxiv.org/abs/2507.20755","https://doi.org/10.1002/aaai.12126"],"source_refs":["AIBD-0099-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The comparison with round-robin calling favored the model, but its p-value was .098. The comparison with messaging alone had p = .044.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Regular messages alone, and regular messages with staff calls allocated in rotation.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Researchers randomly assigned participants to model-selected calls, round-robin calls, or regular messaging alone.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Engagement drops fell by 32.0% versus regular messages alone and 28.3% versus round-robin calling.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Each calling group received about 225 staff calls a week for seven weeks.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"AIBD-0099-E01, AIBD-0099-E02, AIBD-0099-E03","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Call logs identified weeks in which a participant listened to at least one message for more than 30 seconds.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Cumulative weekly drops in listening engagement.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Public SAHELI code and a data dictionary cover the later deployed system. They do not identify the software release used in the 2021 trial.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"The groups contained 7,668, 7,668, and 7,667 participants, respectively.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The study measured listening, not clinical health. Cumulative drops can count the same participant in more than one week.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0099-reported-measure","evaluation_data":null},{"case_id":"AIBD-0041","case_url":"records/AIBD-0041.html","measure":"a reconstruction of 393.1 by 507.4 centimeters","reported_result":"a reconstruction of 393.1 by 507.4 centimeters","baseline":"the remaining original canvas of 379.5 by 436 centimeters","result":"a reconstruction of 393.1 by 507.4 centimeters, with strips of 64.4 cm on the left, 7 cm on the right, 23.3 cm at the top and 11.3 cm at the bottom","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.rijksmuseum.nl/en/press/press-releases/for-the-first-time-in-300-years-the-night-watch-is-complete-again","https://hyperallergic.com/artificial-intelligence-helps-reconstruct-the-night-watch-rembrandt/","https://doi.org/10.54874/9785605605652.2026.11.27"],"source_refs":["AIBD-0041-source-0001","AIBD-0041-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0041-reported-measure","evaluation_data":null},{"case_id":"AIBD-0011","case_url":"records/AIBD-0011.html","measure":"After 4 h 31 min of analysis the system detected him and returned his coordinates","reported_result":"After 4 h 31 min of analysis the system detected him and returned his coordinates","baseline":"Psy zgubiły trop (the dogs lost the scent); police and rescuers could not find him","result":"Po prawie 24 godzinach od zginięcia program wskazał potencjalnego poszukiwanego (after almost 24 hours the program indicated the probable missing person)","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.se.pl/rzeszow/bieszczady-pierwszy-raz-w-polsce-dron-odnalazl-zagubionego-czlowieka-galeria-aa-Lokr-ga7B-yzAt.html","https://doi.org/10.3390/rs13234903","https://oiot.pl/dron-saruav-uratowal-zycie-czlowiekowi-w-polsce/","https://www.swiatdronow.pl/polski-system-saruav-uratowal-zycie-zaginionemu-mezczyznie","https://uwr.edu.pl/en/another-missing-person-found-using-the-saruav-system/","https://web.archive.org/web/2026/https://www.mdpi.com/2071-1050/14/1/327","https://arxiv.org/pdf/2307.14527"],"source_refs":["AIBD-0011-source-0001","AIBD-0011-source-0002","AIBD-0011-source-0003","AIBD-0011-source-0004","AIBD-0011-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the ground search by police officers and rescuers with tracking dogs, which had failed","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: Psy zgubiły trop (the dogs lost the scent); police and rescuers could not find him. Result: Po prawie 24 godzinach od zginięcia program wskazał potencjalnego poszukiwanego (after almost 24 hours the program indicated the probable missing person).","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"After 4 h 31 min of analysis the system detected him and returned his coordinates","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0011-reported-measure","evaluation_data":null},{"case_id":"AIBD-0248","case_url":"records/AIBD-0248.html","measure":null,"reported_result":null,"baseline":"従来システムでは難しかった歩行者やベビーカー、車イスなどを含む\"人\"の検知","result":"AIの画像解析技術が活用されており","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://news.mynavi.jp/article/20200812-1216717","https://optage.co.jp/press/2020/press_34.html","https://oecd.ai/en/incidents/2020-08-11-4fb3","https://optage.co.jp/announce/announce_202115.html","https://www.kobe-np.co.jp/news/keizai/202108/0014560805.shtml"],"source_refs":["AIBD-0248-source-0002","AIBD-0248-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the conventional crossing-detection system (従来システム)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: 従来システムでは難しかった歩行者やベビーカー、車イスなどを含む\"人\"の検知. Result: AIの画像解析技術が活用されており.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0248-reported-measure","evaluation_data":null},{"case_id":"AIBD-0648","case_url":"records/AIBD-0648.html","measure":"Referral uptake within 30 days.","reported_result":"raised 30-day referral attendance to 51.5% from 39.6% with delayed human grading","baseline":"39.6% referral attendance within 30 days","result":"51.5% referral attendance within 30 days","baseline_value":null,"result_value":null,"unit":null,"effect":"51.5% vs 39.6% (P = 0.048).","design":"randomized.","sample":"275 adults with referable findings at four diabetes clinics in and around Kigali, March to July 2021.","uncertainty":"Small sample; P value close to 0.05; authors employed by the AI provider.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC9754978/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13272687/"],"source_refs":["AIBD-0648-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"11.9 percentage points.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Referral advice after human grading 3 to 5 days later.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"51.5% vs 39.6% (P = 0.048).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Immediate AI result for the intervention group.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Presentation at referral services within 30 days.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Referral uptake within 30 days.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Registered as PACTR202101512465690.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"275 adults with referable findings at four diabetes clinics in and around Kigali, March to July 2021.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small sample; P value close to 0.05; authors employed by the AI provider.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0648-reported-measure","evaluation_data":null},{"case_id":"AIBD-0082","case_url":"records/AIBD-0082.html","measure":"prevented at least R$8.33 million of overpriced purchases","reported_result":"prevented at least R$8.33 million of overpriced purchases","baseline":"valor estimado de R$ 34 mi","result":"reduziu o valor total do pregão em R$ 2.381.663,00; all items bought for R$ 14.315.793,42 on 21 July 2021","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/20260412063523id_/https://www.gov.br/cgu/pt-br/assuntos/noticias/2023/10/cgu-atua-em-editais-de-licitacao-em-hospitais-federais-no-rio-de-janeiro-e-evita-desperdicio-de-cerca-de-r-8-33-mi","https://revista.cgu.gov.br/Revista_da_CGU/article/view/530","https://web.archive.org/web/20260207114437id_/https://www.gov.br/cgu/pt-br/assuntos/noticias/2024/12/orgaos-federais-economizaram-em-2024-257-milhoes-de-reais-com-apoio-do-robo-alice-da-controladoria-geral-da-uniao","https://pmc.ncbi.nlm.nih.gov/articles/PMC10039446/"],"source_refs":["AIBD-0082-source-0001","AIBD-0082-source-0002","AIBD-0082-source-0003","AIBD-0082-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the original estimated value of Pregão 15/2021, which reused the price survey of Pregão 05/2021","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: valor estimado de R$ 34 mi. Result: reduziu o valor total do pregão em R$ 2.381.663,00; all items bought for R$ 14.315.793,42 on 21 July 2021.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"prevented at least R$8.33 million of overpriced purchases","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0082-reported-measure","evaluation_data":null},{"case_id":"AIBD-0268","case_url":"records/AIBD-0268.html","measure":"at four months his ejection fraction had improved to 55%","reported_result":"at four months his ejection fraction had improved to 55%","baseline":"The ECG showed sinus tachycardia, premature ventricular complexes with rightward deviation, and a nonspecific T-wave abnormality (echocardiography then showed a left ventricular ejection fraction of 41%)","result":"high probabilities of reduced ejection fraction (97.88% and 98.67%); at a 4-month follow-up visit, the patient had improved left ventricular ejection fraction to 55%","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8890344/","https://doi.org/10.1111/pace.14980","https://pmc.ncbi.nlm.nih.gov/articles/PMC11205191/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12666949/","https://medcitynews.com/2024/04/eko-health-gains-fda-clearance-for-cardiology-algorithm-that-can-detect-heart-failure-in-15-seconds/","https://pmc.ncbi.nlm.nih.gov/articles/PMC8789562/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13676012/","https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm?ID=K232699"],"source_refs":["AIBD-0268-source-0001","AIBD-0268-source-0002","AIBD-0268-source-0003","AIBD-0268-source-0004","AIBD-0268-source-0005","AIBD-0268-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the conventional reading of the same ECGs, which showed only nonspecific abnormalities","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: The ECG showed sinus tachycardia, premature ventricular complexes with rightward deviation, and a nonspecific T-wave abnormality (echocardiography then showed a left ventricular ejection fraction of 41%). Result: high probabilities of reduced ejection fraction (97.88% and 98.67%); at a 4-month follow-up visit, the patient had improved left ventricular ejection fraction to 55%.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"at four months his ejection fraction had improved to 55%","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0268-reported-measure","evaluation_data":null},{"case_id":"AIBD-0354","case_url":"records/AIBD-0354.html","measure":null,"reported_result":null,"baseline":null,"result":"OCHA chose the 510 model after scoring the three models in an evaluation matrix; the annex says that the tight timeline led OCHA to favor a model that could run quickly","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://reliefweb.int/attachments/78d0268e-b476-3a55-ab82-94e7478fc408/Anticipatory%20Action%20Framework%20-%20Philippines%20-%20Typhoons%20-%202021-2022.pdf","https://www.un.org/en/delegate/how-anticipatory-action-shaped-response-philippines%E2%80%99-twin-typhoons","https://reliefweb.int/report/philippines/anticipatory-action-framework-philippines-2021-2022","https://reliefweb.int/report/philippines/after-action-review-cerf-anticipatory-action-typhoons","https://www.unocha.org/publications/report/philippines/philippines-cerf-aa-anticipatory-action-framework-2023-revision","https://reliefweb.int/report/philippines/anticipatory-action-framework-philippines-tropical-cyclones-03-october-2025","https://reliefweb.int/report/philippines/philippines-tropical-cyclones-kalmaegitino-and-fung-wonguwan-flash-update-no1-07-november-9-pm-local-time","https://www.unocha.org/publications/report/philippines/philippines-multiple-tropical-cyclones-humanitarian-needs-and-priorities-response-overview-report-november-2024-april-2025","https://nhess.copernicus.org/articles/24/309/2024/","https://www.anticipation-hub.org/Documents/Framework_documents/OCHA-Philippines-Cyclones-Anticipatory-Action-Framework-2024.pdf","https://reliefweb.int/attachments/4e317214-819a-44d8-99f7-0ca9816dcfc9/OCHA%20AA%20Framework-Philippines%202026-Tropical%20Cyclones%20-%20Endorsed.pdf"],"source_refs":["AIBD-0354-source-0002","AIBD-0354-source-0003","AIBD-0354-source-0004","AIBD-0354-source-0005","AIBD-0354-source-0006","AIBD-0354-source-0007","AIBD-0354-source-0001","AIBD-0354-source-0008","AIBD-0354-source-0009"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0354-reported-measure","evaluation_data":null},{"case_id":"AIBD-0624","case_url":"records/AIBD-0624.html","measure":"Per-slide sensitivity and specificity.","reported_result":"improvement in sensitivity in 7.3% of biopsy images","baseline":"Specificity 88.45%","result":"Specificity 89.50%","baseline_value":null,"result_value":null,"unit":null,"effect":"Sensitivity improved in 7.3% of cancer slides; specificity 88.45% to 89.50%.","design":"quasi-experimental.","sample":"Sixteen pathologists; 610 whole slide images originally, 171 cancer slides analyzed for the sensitivity change.","uncertainty":"Measured per biopsy image; the per-patient gain is likely lower; retrospective images.","comparison_basis":"measured","source_urls":["https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN200080.pdf"],"source_refs":["AIBD-0624-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"7.3% net sensitivity gain per biopsy image.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Unassisted read by the same pathologist.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sensitivity improved in 7.3% of cancer slides; specificity 88.45% to 89.50%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One assisted read per image.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sequential unassisted then assisted reads of whole slide images against diagnostic report ground truth.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Per-slide sensitivity and specificity.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"None public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sixteen pathologists; 610 whole slide images originally, 171 cancer slides analyzed for the sensitivity change.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Measured per biopsy image; the per-patient gain is likely lower; retrospective images.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0624-reported-measure","evaluation_data":null},{"case_id":"AIBD-0110","case_url":"records/AIBD-0110.html","measure":null,"reported_result":null,"baseline":"An unconscious rider on a road with little traffic and few pedestrians","result":"An ambulance summoned and his girlfriend told by the watch","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.zaobao.com.sg/news/singapore/story20210928-1198175","https://mothership.sg/2021/09/motorcyclist-apple-watch-saves-life-ang-mo-kio/","https://support.apple.com/en-us/108896"],"source_refs":["AIBD-0110-source-0001","AIBD-0110-source-0002","AIBD-0110-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0110-reported-measure","evaluation_data":null},{"case_id":"AIBD-0026","case_url":"records/AIBD-0026.html","measure":"detected an irregular heart rate of 164 beats per minute with irregular rhythm","reported_result":"detected an irregular heart rate of 164 beats per minute with irregular rhythm","baseline":"palpitations, sweating, anxiety and mild dizziness for half an hour while at work","result":"he sought a doctor at once after the watch showed an irregular rhythm at 164 beats per minute","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10432923/","https://europepmc.org/article/PMC/PMC10432923","https://www.accessdata.fda.gov/cdrh_docs/pdf20/K200948.pdf","https://web.archive.org/web/20210928163659/https://www.fitbit.com/global/us/technology/ecg"],"source_refs":["AIBD-0026-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"detected an irregular heart rate of 164 beats per minute with irregular rhythm","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"detected an irregular heart rate of 164 beats per minute with irregular rhythm","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0026-reported-measure","evaluation_data":null},{"case_id":"AIBD-0074","case_url":"records/AIBD-0074.html","measure":null,"reported_result":null,"baseline":"have only survived in grainy black-and-white photographs, devoid of the sumptuous, aurous pigment on which the painter's name was made","result":"a digital reconstruction of his Faculty works — in full and glorious color","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://jingdailyculture.com/google-arts-and-culture-klimt-ai/","https://www.belvedere.at/en/gustav-klimt-pigment-pixel","https://www.i-programmer.info/news/105-artificial-intelligence/14928-google-ai-recreates-lost-klimt-artworks-.html","https://artsandculture.google.com/story/01001/SQWxuZfE5ki3mQ","https://www.wired.com/story/artificial-intelligence-reviving-lost-art/"],"source_refs":["AIBD-0074-source-0001","AIBD-0074-source-0002","AIBD-0074-source-0003","AIBD-0074-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the surviving black-and-white photographs of the destroyed paintings","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: have only survived in grainy black-and-white photographs, devoid of the sumptuous, aurous pigment on which the painter's name was made. Result: a digital reconstruction of his Faculty works — in full and glorious color.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0074-reported-measure","evaluation_data":null},{"case_id":"AIBD-0073","case_url":"records/AIBD-0073.html","measure":null,"reported_result":null,"baseline":"impossible for symphony experts to go beyond that first movement","result":"a complete symphony, with third and fourth movements generated with the AI, premiered on 9 October 2021","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://theconversation.com/how-a-team-of-musicologists-and-computer-scientists-completed-beethovens-unfinished-10th-symphony-168160","https://www.beethoven-orchester.de/konzerte/beethoven-x","https://www.euronews.com/video/2021/10/13/artificial-intelligence-helps-complete-beethoven-symphony","https://www.telekom.hu/about_us/press_room/press_releases/2021/october_1","https://www.br-klassik.de/aktuell/news-kritik/kritik-urauffuehrung-beethoven-10-symphonie-kuenstliche-intelligenz-computer-bonn-100.html"],"source_refs":["AIBD-0073-source-0001","AIBD-0073-source-0002","AIBD-0073-source-0003","AIBD-0073-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: earlier human completion attempts from the sketches, such as Barry Cooper's 1988 version of the first movement","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: impossible for symphony experts to go beyond that first movement. Result: a complete symphony, with third and fourth movements generated with the AI, premiered on 9 October 2021.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0073-reported-measure","evaluation_data":null},{"case_id":"AIBD-0043","case_url":"records/AIBD-0043.html","measure":"covered more than 90% of the scaffold, roughly twice the coverage of the previous model","reported_result":"covered more than 90% of the scaffold, roughly twice the coverage of the previous model","baseline":"16 Nups and 35 MDa (46% of the scaffold) of the previous model","result":"25 out of ~30 Nups, and account for 70 MDa of the molecular weight of the NPC (>90% of the scaffold molecular weight)","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.biorxiv.org/content/10.1101/2021.10.26.465776v1.full","https://doi.org/10.1126/science.abm9506","https://deepmind.google/science/alphafold/","https://zenodo.org/records/5068265","https://pubmed.ncbi.nlm.nih.gov/36641895/","https://pmc.ncbi.nlm.nih.gov/articles/PMC10054137/"],"source_refs":["AIBD-0043-source-0001","AIBD-0043-source-0002","AIBD-0043-source-0003","AIBD-0043-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the previous structural model of the human NPC scaffold","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 16 Nups and 35 MDa (46% of the scaffold) of the previous model. Result: 25 out of ~30 Nups, and account for 70 MDa of the molecular weight of the NPC (>90% of the scaffold molecular weight).","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"covers more than 90% of the NPC scaffold, roughly doubling previous structural coverage","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0043-reported-measure","evaluation_data":null},{"case_id":"AIBD-0339","case_url":"records/AIBD-0339.html","measure":null,"reported_result":null,"baseline":null,"result":"No measured change; the case studies list expected benefits only","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://htn.co.uk/2021/09/20/video-kettering-general-hospital-opens-htn-now-focusing-on-ai-patient-flow-and-bed-management/","https://www.gov.uk/government/case-studies/using-ai-to-improve-patient-bed-allocation-decisions-in-hospital","https://web.archive.org/web/20231123150115/https://transform.england.nhs.uk/ai-lab/explore-all-resources/develop-ai/improving-hospital-bed-allocation-using-ai/","https://github.com/nhsx/skunkworks-bed-allocation"],"source_refs":["AIBD-0339-source-0002","AIBD-0339-source-0003","AIBD-0339-source-0001","AIBD-0339-source-0004","AIBD-0339-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0339-reported-measure","evaluation_data":null},{"case_id":"AIBD-0365","case_url":"records/AIBD-0365.html","measure":"avoided a loss of 190,000 yuan","reported_result":"avoided a loss of 190,000 yuan","baseline":"a 190,000 yuan transfer about to be sent","result":"the payment canceled and the loss avoided","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.chinaz.com/2021/1214/1341135.shtml","https://mobile.zol.com.cn/783/7832465.html","http://newsxmwb.xinmin.cn/shizheng/szt/2021/12/17/32082134.html","https://oecd.ai/en/incidents/2021-12-15-3cd7","https://finance.sina.com.cn/tech/2021-12-14/doc-ikyamrmy8901677.shtml","http://www.news.cn/tech/20220922/64f733504fa640c894b990ba2e31142c/c.html"],"source_refs":["AIBD-0365-source-0001","AIBD-0365-source-0002","AIBD-0365-source-0003","AIBD-0365-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"avoided a loss of 190,000 yuan","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"avoided a loss of 190,000 yuan","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0365-reported-measure","evaluation_data":null},{"case_id":"AIBD-0050","case_url":"records/AIBD-0050.html","measure":null,"reported_result":null,"baseline":null,"result":"significant improvements in the Assessment of Capacity for Myoelectric Control (ACMC) outcome measure while using pattern recognition control compared to direct control","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://europepmc.org/abstract/MED/36355739","https://europepmc.org/abstract/MED/34910710","https://pmc.ncbi.nlm.nih.gov/articles/PMC6331198/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11365260/"],"source_refs":["AIBD-0050-source-0001","AIBD-0050-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"direct myoelectric control","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Result: significant improvements in the Assessment of Capacity for Myoelectric Control (ACMC) outcome measure while using pattern recognition control compared to direct control.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0050-reported-measure","evaluation_data":null},{"case_id":"AIBD-0084","case_url":"records/AIBD-0084.html","measure":"more than 20,000 pools being taxed for 2022, close to €10 million of extra revenue","reported_result":"more than 20,000 pools being taxed for 2022, close to €10 million of extra revenue","baseline":"the pools were not declared and not taxed","result":"20,000 undeclared pools identified and 10 million euros collected for the communes concerned, according to the tax administration","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.ccomptes.fr/sites/default/files/2023-11/20231115-Detection-fraude-fiscale-des-particuliers.pdf","https://www.impots.gouv.fr/actualite/generalisation-du-foncier-innovant","https://www.collectivites-locales.gouv.fr/animer-les-territoires/environnement-et-urbanisme/le-cadastre/le-foncier-innovant","https://www.impots.gouv.fr/actualite/1-les-grandes-lignes-du-projet-du-foncier-innovant-afin-de-garantir-une-meilleure","https://www.assemblee-nationale.fr/dyn/15/questions/QANR5L15QE41677.pdf","https://www.senat.fr/rap/r23-491/r23-4911.pdf"],"source_refs":["AIBD-0084-source-0001","AIBD-0084-source-0002","AIBD-0084-source-0003","AIBD-0084-source-0004","AIBD-0084-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"more than 20,000 pools being taxed for 2022, close to €10 million of extra revenue","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"more than 20,000 pools being taxed for 2022, close to €10 million of extra revenue","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0084-reported-measure","evaluation_data":null},{"case_id":"AIBD-0096","case_url":"records/AIBD-0096.html","measure":"2,337 TB cases diagnosed, 97% treated","reported_result":"2,337 TB cases diagnosed, 97% treated","baseline":"NNS 205.6 (public facilities), 179.0 (private facilities), 127.8 (medicine vendors and community pharmacists), 49.0 (contact investigation), 388.5 (community active case-finding)","result":"NNS 44.9 with PDX-CAD","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11841120/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12257310/","https://kncvnigeria.org/wp-content/uploads/2026/03/2026-KNCV-Nigeria-Annual-Report.pdf"],"source_refs":["AIBD-0096-source-0001","AIBD-0096-source-0002","AIBD-0096-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"KNCV Nigeria's other 2022 case-finding strategies in the same states without portable X-ray (intensified case finding in public and private facilities, medicine vendors and community pharmacists, contact investigation, community active case finding)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: NNS 205.6 (public facilities), 179.0 (private facilities), 127.8 (medicine vendors and community pharmacists), 49.0 (contact investigation), 388.5 (community active case-finding). Result: NNS 44.9 with PDX-CAD.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"2,337 TB cases diagnosed, 97% treated","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0096-reported-measure","evaluation_data":null},{"case_id":"AIBD-0199","case_url":"records/AIBD-0199.html","measure":"28.4% early-postemergence and 36.5% mid-postemergence herbicide savings","reported_result":"28.4% early-postemergence and 36.5% mid-postemergence herbicide savings","baseline":"full broadcast applications of the postemergence herbicides (traditional broadcast program)","result":"Targeted applications saved 28.4% in early-postemergence and 36.5% in mid-postemergence treatments, with comparable weed control","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.cambridge.org/core/journals/weed-technology/article/comparing-herbicide-application-methods-with-see-spray-technology-in-soybean/05967C11D1BF0B174A13BE687B865AF2","https://doi.org/10.1017/wet.2024.70","https://www.globalagtechinitiative.com/in-field-technologies/robotics-automation/john-deere-customers-use-see-spray-technology-across-five-million-acres-in-2025/","https://doi.org/10.1017/wet.2025.10056"],"source_refs":["AIBD-0199-source-0001","AIBD-0199-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"traditional broadcast herbicide applications in the same randomized trials","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: full broadcast applications of the postemergence herbicides (traditional broadcast program). Result: Targeted applications saved 28.4% in early-postemergence and 36.5% in mid-postemergence treatments, with comparable weed control.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"28.4% early-postemergence and 36.5% mid-postemergence herbicide savings","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0199-reported-measure","evaluation_data":null},{"case_id":"AIBD-0487","case_url":"records/AIBD-0487.html","measure":null,"reported_result":null,"baseline":"riders and fleet operators without affordable credit for electric two-wheelers","result":"more than 20,000 electric two-wheelers financed since 2022, by the developer's count","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://hubbis.com/news/aizen-global-to-expand-its-e-mobility-banking-service-in-indonesia","https://web.archive.org/web/2026/https://technode.global/2023/11/10/aizen-global-and-grab-forge-partnership-in-mobility-finance/","https://www.hankyung.com/article/2025072405351"],"source_refs":["AIBD-0487-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0487-reported-measure","evaluation_data":null},{"case_id":"AIBD-0290","case_url":"records/AIBD-0290.html","measure":"The 35 mm tumor was staged pT2a pN0 M0 with clear margins","reported_result":"The 35 mm tumor was staged pT2a pN0 M0 with clear margins","baseline":"his lung adenocarcinoma would have continued to grow undetected, curative surgery may not have been an option","result":"The upper lobe adenocarcinoma measured 35 mm and was staged as pT2a pN0 M0 with clear resection margins.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8733372/"],"source_refs":["AIBD-0290-source-0001","AIBD-0290-source-0002","AIBD-0290-source-0003","AIBD-0290-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: no smartwatch (the authors' counterfactual for the same patient)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: his lung adenocarcinoma would have continued to grow undetected, curative surgery may not have been an option. Result: The upper lobe adenocarcinoma measured 35 mm and was staged as pT2a pN0 M0 with clear resection margins.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"found a 28 mm lung adenocarcinoma, resected","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0290-reported-measure","evaluation_data":null},{"case_id":"AIBD-0111","case_url":"records/AIBD-0111.html","measure":"Hürriyet put the lead at about 10 seconds","reported_result":"Hürriyet put the lead at about 10 seconds","baseline":"No warning before the shaking","result":"A phone alert about 10 seconds before the shaking, by Hürriyet's account; other outlets gave up to one minute","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.hurriyet.com.tr/gundem/erken-uyari-sistemi-tartismasi-depremin-olacagini-10-saniye-onceden-bilmek-guvenli-mi-41976168","https://blog.google/products/android/earthquake-detection-and-alerts/","https://www.yenisafak.com/teknoloji/google-android-deprem-uyari-sistemi-nasil-acilir-deprem-uyari-sistemi-nedir-deprem-uyari-sistemi-hangi-telefonlarda-acilir-3728848","https://www.donanimhaber.com/google-erken-uyari-sistemi-alanya-depremini-onceden-bildi--143452","https://www.hurriyet.com.tr/teknoloji/google-android-deprem-uyari-sistemini-tanitti-41995996","https://indigodergisi.com/2022/01/google-deprem-uyari-sistemi-nasil-ailir/","https://www.ntv.com.tr/turkiye/deprem-uyari-sistemi-nedir-nasil-calisir,PtR3VF38DUOdDzApupgnQQ","https://www.yeniakit.com.tr/haber/android-deprem-uyari-sistemi-nedir-nasil-yapilir-nasil-kullanilir-1618101.html","https://www.hurriyet.com.tr/galeri-deprem-uyari-sistemi-nasil-yapilir-nedir-android-erken-deprem-uyari-sistemi-kullanim-detaylari-41984292","https://earthquake.usgs.gov/earthquakes/eventpage/us7000g94d","https://www.zdnet.com/article/android-opens-up-earthquake-alerts-and-end-to-end-encrypted-messages/","https://blog.google/products-and-platforms/platforms/android/new-features-summer-2021/","https://crisisresilience.google/android-early-earthquake-warnings","https://doi.org/10.1126/science.ads4779","https://www.bbc.com/news/technology-66316462"],"source_refs":["AIBD-0111-source-0002","AIBD-0111-source-0003","AIBD-0111-source-0004","AIBD-0111-source-0005","AIBD-0111-source-0006","AIBD-0111-source-0007","AIBD-0111-source-0008","AIBD-0111-source-0009","AIBD-0111-source-0010","AIBD-0111-source-0011","AIBD-0111-source-0012","AIBD-0111-source-0001","AIBD-0111-source-0013","AIBD-0111-source-0014"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Hürriyet put the lead at about 10 seconds","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Hürriyet put the lead at about 10 seconds","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0111-reported-measure","evaluation_data":null},{"case_id":"AIBD-0289","case_url":"records/AIBD-0289.html","measure":"by the next day her heart rate was 60–70 bpm","reported_result":"by the next day her heart rate was 60–70 bpm","baseline":"a potential adverse event","result":"Early recognition of bradycardia","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8793425/"],"source_refs":["AIBD-0289-source-0001","AIBD-0289-source-0002","AIBD-0289-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: presentation without the smartwatch alert","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: a potential adverse event. Result: Early recognition of bradycardia.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"by the next day her heart rate was 60–70 bpm","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0289-reported-measure","evaluation_data":null},{"case_id":"AIBD-0037","case_url":"records/AIBD-0037.html","measure":null,"reported_result":null,"baseline":"At four years of age, he was clinically diagnosed with Cornelia de Lange syndrome (CdLS; OMIM 122470, 300590, 300882, 610759, and 614701) due to his facial dysmorphism, microcephaly, short stature, and intellectual disability.","result":"Face2Gene introduced the following candidate diseases: Angelman syndrome (AS, OMIM 105833), Witteveen-Kolk syndrome (WKOS; MIM 613406), Rubinstein-Taybi syndrome (RTS; OMIM 180849), and MRD7.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC9822525/","https://web.archive.org/web/20250206201919/https://www.ncbi.nlm.nih.gov/clinvar/variation/1691094/"],"source_refs":["AIBD-0037-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the clinical diagnosis of Cornelia de Lange syndrome made at age four","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: At four years of age, he was clinically diagnosed with Cornelia de Lange syndrome (CdLS; OMIM 122470, 300590, 300882, 610759, and 614701) due to his facial dysmorphism, microcephaly, short stature, and intellectual disability. Result: Face2Gene introduced the following candidate diseases: Angelman syndrome (AS, OMIM 105833), Witteveen-Kolk syndrome (WKOS; MIM 613406), Rubinstein-Taybi syndrome (RTS; OMIM 180849), and MRD7.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0037-reported-measure","evaluation_data":null},{"case_id":"AIBD-0519","case_url":"records/AIBD-0519.html","measure":"Information retained in the signing and comprehension by Deaf viewers.","reported_result":"A significant amount of key information was lost or distorted, and Deaf viewers had difficulty understanding the signing.","baseline":null,"result":"Key information lost or distorted in the avatar signing","baseline_value":null,"result_value":null,"unit":null,"effect":"A significant amount of key information was lost or distorted, and Deaf viewers had difficulty understanding the signing.","design":"qualitative evaluation.","sample":"Sign language avatars on several major Chinese TV channels during the Beijing 2022 Winter Olympics; sample size not reported.","uncertainty":"Reported in an opinion column with no published methods or sample; covers several broadcasters' avatars without separating Baidu's; the author once consulted on an avatar for an unnamed major Chinese tech company.","comparison_basis":"measured","source_urls":["https://www.sixthtone.com/news/1017316","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-830","https://global.chinadaily.com.cn/a/202202/08/WS6201cb56a310cdd39bc85425.html"],"source_refs":["AIBD-0519-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The original spoken audio.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"qualitative evaluation.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"A significant amount of key information was lost or distorted, and Deaf viewers had difficulty understanding the signing.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"qualitative evaluation.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Transcription and back-translation of avatar signing compared with the original audio, plus interviews with Deaf users and sign language professionals.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Information retained in the signing and comprehension by Deaf viewers.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sign language avatars on several major Chinese TV channels during the Beijing 2022 Winter Olympics; sample size not reported.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Reported in an opinion column with no published methods or sample; covers several broadcasters' avatars without separating Baidu's; the author once consulted on an avatar for an unnamed major Chinese tech company.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0519-reported-measure","evaluation_data":null},{"case_id":"AIBD-0164","case_url":"records/AIBD-0164.html","measure":"narrowed the separation to 6.6 cm with a combined root-mean-square error of 3.7 cm","reported_result":"narrowed the separation to 6.6 cm with a combined root-mean-square error of 3.7 cm","baseline":"the authors judge that existing approaches could probably stabilize droplets, but only with a large investment of design work","result":"droplets held over the whole 200 ms control window from a reward function and an adjusted handover state","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.nature.com/articles/s41586-021-04301-9","https://www.technologyreview.com/2022/02/16/1045470/deepminds-ai-can-control-superheated-plasma-inside-a-fusion-reactor/","https://arxiv.org/abs/2307.11546"],"source_refs":["AIBD-0164-source-0001","AIBD-0164-source-0002","AIBD-0164-source-0003","AIBD-0164-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"narrowed the separation to 6.6 cm with a combined RMSE of 3.7 cm","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"narrowed the separation to 6.6 cm with a combined RMSE of 3.7 cm","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0164-reported-measure","evaluation_data":null},{"case_id":"AIBD-0649","case_url":"records/AIBD-0649.html","measure":"Diagnostic accuracy.","reported_result":"diagnosed 53.9% correctly against 43.8% without it","baseline":"43.8% diagnostic accuracy","result":"53.9% diagnostic accuracy","baseline_value":null,"result_value":null,"unit":null,"effect":"53.9% vs 43.8% overall; 54.7% vs 30.7% for non-dermatology trainees.","design":"randomized.","sample":"576 consecutive adults with suspicious skin lesions at a tertiary hospital in Seoul; eight non-expert physicians.","uncertainty":"Unmasked single-center design; no significant gain for dermatology residents; the developer is the first author.","comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/35183551/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11988623/","https://b2020.modelderm.com/"],"source_refs":["AIBD-0649-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"10.1 percentage points overall.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Unaided examination.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"53.9% vs 43.8% overall; 54.7% vs 30.7% for non-dermatology trainees.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Real-time algorithm assistance in the AI group.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Top-1 diagnostic accuracy against final diagnosis.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Diagnostic accuracy.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Algorithm available at b2020.modelderm.com.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"576 consecutive adults with suspicious skin lesions at a tertiary hospital in Seoul; eight non-expert physicians.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Unmasked single-center design; no significant gain for dermatology residents; the developer is the first author.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0649-reported-measure","evaluation_data":null},{"case_id":"AIBD-0581","case_url":"records/AIBD-0581.html","measure":null,"reported_result":null,"baseline":"half an hour or longer","result":"just seconds","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://web.archive.org/web/20260711183830/https://mayomagazine.mayoclinic.org/2025/01/ai-in-stroke-care/","https://mayomagazine.mayoclinic.org/2025/01/ai-in-stroke-care/","https://web.archive.org/web/20260520215655/https://newsnetwork.mayoclinic.org/discussion/florida-mom-reunites-with-icu-team-after-stroke/"],"source_refs":["AIBD-0581-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0581-reported-measure","evaluation_data":null},{"case_id":"AIBD-0610","case_url":"records/AIBD-0610.html","measure":"Door-to-groin time.","reported_result":"door-to-groin time fell by 11.2 minutes","baseline":"Median door-to-groin time of 100 minutes before exposure","result":"11.2-minute reduction in door-to-groin time","baseline_value":null,"result_value":null,"unit":null,"effect":"-11.2 minutes (95% CI -18.22 to -4.2).","design":"randomized.","sample":"243 patients with large vessel occlusion treated with thrombectomy at four Houston comprehensive stroke centers.","uncertainty":"Transfers and in-hospital strokes excluded; no difference in 90-day functional independence; senior author funded by the maker.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10507590/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13475654/"],"source_refs":["AIBD-0610-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"11.2 minutes faster door-to-groin.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Unexposed periods in the stepped-wedge design.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"-11.2 minutes (95% CI -18.22 to -4.2).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"All emergency department stroke imaging after activation.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Treatment timestamps from stroke registry data; 90-day modified Rankin Scale.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Door-to-groin time.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"ClinicalTrials.gov NCT05838456; supplement.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"243 patients with large vessel occlusion treated with thrombectomy at four Houston comprehensive stroke centers.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Transfers and in-hospital strokes excluded; no difference in 90-day functional independence; senior author funded by the maker.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0610-reported-measure","evaluation_data":null},{"case_id":"AIBD-0637","case_url":"records/AIBD-0637.html","measure":"Study 1: Door-to-groin time. Study 2: Door-to-groin time and other workflow times (standardized mean difference).","reported_result":"fell by 11.2 minutes (95% CI 4.2 to 18.2) after activation","baseline":"Median door-to-groin time of 100 minutes before activation","result":"Door-to-groin time 11.2 minutes shorter after activation (adjusted)","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: 11.2 minutes shorter (95% CI 4.2 to 18.2). Study 2: Door-to-groin SMD -0.50 (95% CI -0.66 to -0.35).","design":"Study 1: randomized. Study 2: other.","sample":"Study 1: 243 emergency department patients with large vessel occlusion stroke treated with thrombectomy at four centers, January 2021 to February 2022. Study 2: 12 studies, 15,595 patients, stroke centers in several countries.","uncertainty":"Study 1: Small number of clusters; transfers excluded; senior author received grants from the developer. Study 2: Mostly observational source studies, several by developer-linked authors.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10507590/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12596299/"],"source_refs":["AIBD-0637-source-0001","AIBD-0637-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 11.2 minutes. Study 2: Shorter workflow times; no significant change in clinical outcomes.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Unexposed periods at the same centers. Study 2: Periods before Viz.ai implementation.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: randomized. Study 2: other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 11.2 minutes shorter (95% CI 4.2 to 18.2). Study 2: Door-to-groin SMD -0.50 (95% CI -0.66 to -0.35).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Patients treated after activation at each center. Study 2: Patients treated after implementation.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: randomized. Study 2: other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Door-to-groin time from stroke registry data; mixed-effects linear regression with center as random effect. Study 2: Systematic review and meta-analysis of before/after cohorts and one randomized trial.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Door-to-groin time. Study 2: Door-to-groin time and other workflow times (standardized mean difference).","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Data sharing statement in supplement. Study 2: Supplementary file with extracted data.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 243 emergency department patients with large vessel occlusion stroke treated with thrombectomy at four centers, January 2021 to February 2022. Study 2: 12 studies, 15,595 patients, stroke centers in several countries.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Small number of clusters; transfers excluded; senior author received grants from the developer. Study 2: Mostly observational source studies, several by developer-linked authors.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0637-reported-measure","evaluation_data":null},{"case_id":"AIBD-0027","case_url":"records/AIBD-0027.html","measure":"resting heart rate had risen persistently from about 55 to about 75 beats per minute","reported_result":"resting heart rate had risen persistently from about 55 to about 75 beats per minute","baseline":"Two months before referral, her general practitioner's chest CT was read as normal and the mild thickening of the aortic wall was missed.","result":"The trend in resting heart rate led her to seek care, and the authors report that it contributed to an earlier diagnosis.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC8886383/"],"source_refs":["AIBD-0027-source-0001","AIBD-0027-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"resting heart rate had risen persistently from about 55 to about 75 beats per minute","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"resting heart rate had risen persistently from about 55 to about 75 beats per minute","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0027-reported-measure","evaluation_data":null},{"case_id":"AIBD-0179","case_url":"records/AIBD-0179.html","measure":"about 5 million text messages had been flagged to users as potential scams","reported_result":"about 5 million text messages had been flagged to users as potential scams","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.mha.gov.sg/media-room/newsroom/oral-reply-to-pq-on-effectiveness-of-scamshield-in-preventing-scams/","https://govinsider.asia/intl-en/article/meet-singapores-latest-weapon-against-scams-open-government-products-scamshield","https://www.mddi.gov.sg/other-pages/anti-scam-efforts/","https://www.straitstimes.com/singapore/politics/180-million-scam-calls-blocked-in-first-half-of-2023-scamshield-blocked-80000-phone-numbers"],"source_refs":["AIBD-0179-source-0001","AIBD-0179-source-0002","AIBD-0179-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"about 5 million SMSes had been flagged to users as potential scams","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"about 5 million SMSes had been flagged to users as potential scams","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0179-reported-measure","evaluation_data":null},{"case_id":"AIBD-0351","case_url":"records/AIBD-0351.html","measure":"variable cost per person enrolled; time from registration to enrollment; targeting error.","reported_result":"the hybrid approach cost about half as much per added enrollee as in-person surveys alone","baseline":"the variable cost per person enrolled with in-person proxy-means tests only","result":"around 50% of that variable cost for the hybrid of phone-data targeting and in-person surveys","baseline_value":null,"result_value":null,"unit":null,"effect":"variable cost per enrollee about 50% of in-person only; enrollment within 56 hours of ID validation; targeting error 17% for tele-targeting against 13% for the in-person test.","design":"case study.","sample":"Canva-funded GiveDirectly pilot in TA Khongoni, rural Lilongwe, Malawi; 12,800 recipients in 2022, 3,052 chosen by tele-targeting; about 19,400 USSD registrants.","uncertainty":"Data come from the deployer's internal synthesis; no confidence intervals; about 35.8% of households owned a phone and phone owners were about 20% wealthier; high fixed cost for data agreements and ground-truth surveys.","comparison_basis":"measured","source_urls":["https://documents1.worldbank.org/curated/en/099050724145524418/pdf/P17191311f545a0971b3db17f9d6820d240.pdf","https://www.givedirectly.org/mobileaid","https://www.givedirectly.org/explain_ai","https://www.givedirectly.org/ai-framework","https://www.givedirectly.org/canva/"],"source_refs":["AIBD-0351-source-0005"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The model cut enrollment cost and time but was less accurate than the in-person test and only slightly better than random choice at the 80% cutoff.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"in-person proxy-means-test enrollment.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"variable cost per enrollee about 50% of in-person only; enrollment within 56 hours of ID validation; targeting error 17% for tele-targeting against 13% for the in-person test.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One enrollment round in Phase 1 of the pilot, 2022.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Comparison of three targeting methods (saturation, in-person proxy-means test, tele-targeting plus in-person survey) applied in different areas, with model accuracy checked against a 2,000-household consumption module; figures from GiveDirectly's 2023 internal synthesis.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"variable cost per person enrolled; time from registration to enrollment; targeting error.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"GiveDirectly 2023 internal slides 'Internal Learning Synthesis: Malawi Khongoni Pilot-Teletargeting Retrospective' (not public).","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Canva-funded GiveDirectly pilot in TA Khongoni, rural Lilongwe, Malawi; 12,800 recipients in 2022, 3,052 chosen by tele-targeting; about 19,400 USSD registrants.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Data come from the deployer's internal synthesis; no confidence intervals; about 35.8% of households owned a phone and phone owners were about 20% wealthier; high fixed cost for data agreements and ground-truth surveys.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0351-reported-measure","evaluation_data":null},{"case_id":"AIBD-0017","case_url":"records/AIBD-0017.html","measure":"Study 1: Top-1 restoration accuracy. Study 2: Top-1 restoration accuracy.","reported_result":"raised historians' accuracy from 25 to 72 percent when they worked with it","baseline":"25%","result":"72% (historians using Ithaca); 62% for Ithaca alone","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: Top-1 accuracy rose from 25 percent to 72 percent when the evaluators used Ithaca; Ithaca alone reached 62 percent. Study 2: Ithaca reached 60.9 percent top-1 accuracy on the shared set and 46.5 percent on the new inscriptions; the fine-tuned model reached 61.4 and 48.9 percent.","design":"Study 1: controlled observational. Study 2: other.","sample":"Study 1: Inscriptions held out from the I.PHI corpus; two graduate students of ancient history restored a subset with and without Ithaca's top 20 suggestions. Study 2: A test set shared with the Ithaca paper (PHI identifiers ending in 3) and 4,111 samples drawn from six inscriptions that were not in PHI or online as of December 2023.","uncertainty":"Study 1: Two evaluators only. The authors write that the setup is not a direct comparison of historians and the model and that the evaluators do not stand for all historians. They also write that Ithaca cannot make choices on the basis of historical and material context. Study 2: The new-inscription set comes from six inscriptions only. The author retrained his model on a different split for part of the comparison.","comparison_basis":"measured","source_urls":["https://europepmc.org/article/PMC/PMC8907065","https://www.nature.com/articles/s41586-022-04448-z","https://www.unive.it/pag/16584/?tx_news_pi1%5Bnews%5D=12037","https://www.nature.com/articles/d41586-022-00641-2","https://www.technologyreview.com/2025/07/23/1120574/deepmind-ai-aeneas-helps-historians-interpret-latin-inscriptions/","https://arxiv.org/pdf/2409.13870"],"source_refs":["AIBD-0017-source-0001","AIBD-0017-source-0005"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 47 percentage points of top-1 accuracy for the evaluators.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: The same two evaluators working alone with access to the training set for parallels. Study 2: A Llama 3.1 8B model fine-tuned with instructions on the same task.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: controlled observational. Study 2: other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Top-1 accuracy rose from 25 percent to 72 percent when the evaluators used Ithaca; Ithaca alone reached 62 percent. Study 2: Ithaca reached 60.9 percent top-1 accuracy on the shared set and 46.5 percent on the new inscriptions; the fine-tuned model reached 61.4 and 48.9 percent.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Ranked list of Ithaca's top 20 restoration hypotheses for each damaged text.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: controlled observational. Study 2: other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Character error rate and top-1 accuracy for restoring 1 to 10 hidden characters in held-out inscriptions, with scores averaged for each sequence length. Study 2: Character error rate and top-1 and top-20 accuracy for restoring up to 10 characters, with Ithaca rerun by the author.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Top-1 restoration accuracy. Study 2: Top-1 restoration accuracy.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Code at github.com/deepmind/ithaca under Apache License 2.0, with trained weights and a public interface.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Inscriptions held out from the I.PHI corpus; two graduate students of ancient history restored a subset with and without Ithaca's top 20 suggestions. Study 2: A test set shared with the Ithaca paper (PHI identifiers ending in 3) and 4,111 samples drawn from six inscriptions that were not in PHI or online as of December 2023.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Two evaluators only. The authors write that the setup is not a direct comparison of historians and the model and that the evaluators do not stand for all historians. They also write that Ithaca cannot make choices on the basis of historical and material context. Study 2: The new-inscription set comes from six inscriptions only. The author retrained his model on a different split for part of the comparison.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0017-reported-measure","evaluation_data":null},{"case_id":"AIBD-0366","case_url":"records/AIBD-0366.html","measure":"Seizure frequency reduction from baseline.","reported_result":"Mean reduction of 61.1% in seven patients followed over three months (P = 0.0156); one patient became seizure-free.","baseline":"baseline monthly seizure frequency","result":"mean seizure reduction of 61.1% in seven patients followed over three months","baseline_value":null,"result_value":null,"unit":null,"effect":"Mean reduction of 61.1% in seven patients followed over three months (P = 0.0156); one patient became seizure-free.","design":"before/after.","sample":"10 patients with drug-resistant focal epilepsy at Xuanwu Hospital (3), Second Affiliated Hospital of Zhejiang University (6) and Huashan Hospital (1), enrolled March to November 2022.","uncertainty":"Small sample, short follow-up, no control group, and authors from the implanting teams; no individual result for the first Xuanwu patient.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11188855/","https://www.cnbeta.com.tw/articles/tech/1246089.htm","https://news.mydrivers.com/1/819/819786.htm","https://tech.ifeng.com/c/8EJq2tqI1TE","https://oecd.ai/en/incidents/2022-03-11-af88","https://pmc.ncbi.nlm.nih.gov/articles/PMC11845213/"],"source_refs":["AIBD-0366-source-0006"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Each patient's own baseline seizure frequency.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean reduction of 61.1% in seven patients followed over three months (P = 0.0156); one patient became seizure-free.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean follow-up of 5.1 months.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Monthly seizure counts compared with the mean of the three months before implantation; the first month after stimulation began served to tune settings, and efficacy counts began in the second month.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Seizure frequency reduction from baseline.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"10 patients with drug-resistant focal epilepsy at Xuanwu Hospital (3), Second Affiliated Hospital of Zhejiang University (6) and Huashan Hospital (1), enrolled March to November 2022.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small sample, short follow-up, no control group, and authors from the implanting teams; no individual result for the first Xuanwu patient.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0366-reported-measure","evaluation_data":null},{"case_id":"AIBD-0660","case_url":"records/AIBD-0660.html","measure":"Structure solved or not.","reported_result":"Solved with AlphaFold and RoseTTAFold models after two years of failure.","baseline":"Structure unsolved after two years","result":"Structure solved by molecular replacement with the predicted models","baseline_value":null,"result_value":null,"unit":null,"effect":"Solved with AlphaFold and RoseTTAFold models after two years of failure.","design":"case study.","sample":"One protein crystal data set at 2.45 Å.","uncertainty":"One case in one laboratory.","comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/35362474/"],"source_refs":["AIBD-0660-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Made a previously failed structure solvable.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Isomorphous replacement, anomalous diffraction and molecular replacement with other models.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Solved with AlphaFold and RoseTTAFold models after two years of failure.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Laboratory use.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Success of molecular replacement with different search models.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Structure solved or not.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Methods described in the paper.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One protein crystal data set at 2.45 Å.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One case in one laboratory.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0660-reported-measure","evaluation_data":null},{"case_id":"AIBD-0112","case_url":"records/AIBD-0112.html","measure":"133 packets, 52.5 kg, of factory-made explosive","reported_result":"133 packets, 52.5 kg, of factory-made explosive","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.hurriyet.com.tr/gundem/son-dakika-bakan-soyludan-onemli-aciklamalar-42027018","https://www.cumhuriyet.com.tr/siyaset/goruntuler-ortaya-cikti-aracinda-52-kilo-patlayici-cikan-polis-tutuklandi-1918432","https://www.sondakika.com/haber/haber-polis-memurunun-52-kilo-patlayiciyla-yakalanmasi-14811802","https://www.ntv.com.tr/turkiye/polis-memuru-52-5-kilo-patlayiciyla-yakalandi,FnN59Nq0G0GEzin9tpURRQ","https://www.duvarenglish.com/opposition-mp-ahmet-sik-inquires-about-police-officer-caught-with-52kg-of-explosives-in-hakkari-news-60703","https://www.aa.com.tr/tr/gundem/yapay-zeka-asena-uyusturucunun-izini-suruyor/2603769","https://www.dha.com.tr/yerel-haberler/hakkari/52-kilo-patlayiciyla-yakalanan-polis-memuru-ih-2208374","https://kisadalga.net/haber/detay/aracinda-52-kilo-patlayici-bulunan-polise-21-yil-hapis-cezasi_86842"],"source_refs":["AIBD-0112-source-0002","AIBD-0112-source-0001","AIBD-0112-source-0003","AIBD-0112-source-0004","AIBD-0112-source-0005","AIBD-0112-source-0006","AIBD-0112-source-0007","AIBD-0112-source-0008"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"133 packets, 52.5 kg, of factory-made explosive","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"133 packets, 52.5 kg, of factory-made explosive","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0112-reported-measure","evaluation_data":null},{"case_id":"AIBD-0317","case_url":"records/AIBD-0317.html","measure":"fraudsters were arrested while attempting to withdraw USD 500,000 (KSh 58 million)","reported_result":"fraudsters were arrested while attempting to withdraw USD 500,000 (KSh 58 million)","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.vyntra.com/fraud-cases/","https://fintech.global/2025/06/19/vyntra-launched-by-netguardians-and-intix-to-boost-banking-intelligence/","https://147203409.hs-sites-eu1.com/hubfs/Resources/PDF/Vyntra_Case_study_Sidian_Bank_Kenya.pdf"],"source_refs":["AIBD-0317-source-0001","AIBD-0317-source-0002","AIBD-0317-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"fraudsters were arrested while attempting to withdraw USD 500,000 (KSh 58 million)","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"fraudsters were arrested while attempting to withdraw USD 500,000 (KSh 58 million)","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0317-reported-measure","evaluation_data":null},{"case_id":"AIBD-0045","case_url":"records/AIBD-0045.html","measure":"Percent of PET converted to monomers after 24 hours.","reported_result":"degraded 51 post-consumer PET products in a week","baseline":null,"result":"superior PET-hydrolytic activity relative to both wild-type and engineered alternatives between 30 and 50 °C and a range of pH levels","baseline_value":null,"result_value":null,"unit":null,"effect":"FAST-PETase converted 18.1 percent of the PET in 24 hours, against 97.3 percent for LCC-ICCG.","design":"other.","sample":"Post-consumer colored-flake PET waste powder at 16.5 percent weight in a bioreactor.","uncertainty":"Five authors work for Carbios, which develops LCC-ICCG; each enzyme ran near its own working temperature; one reaction setup.","comparison_basis":"measured","source_urls":["https://www.nature.com/articles/s41586-022-04599-z","https://pmc.ncbi.nlm.nih.gov/articles/PMC11984091/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11420662/","https://pmc.ncbi.nlm.nih.gov/articles/PMC10594578/"],"source_refs":["AIBD-0045-source-0004"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not measured.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"LCC-ICCG, HotPETase and PES-H1 L92F/Q94Y.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"FAST-PETase converted 18.1 percent of the PET in 24 hours, against 97.3 percent for LCC-ICCG.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"24 hours.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Laboratory comparison of four enzymes under one standardized protocol; conversion measured by base consumption and by residual dry weight.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Percent of PET converted to monomers after 24 hours.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Post-consumer colored-flake PET waste powder at 16.5 percent weight in a bioreactor.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Five authors work for Carbios, which develops LCC-ICCG; each enzyme ran near its own working temperature; one reaction setup.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0045-reported-measure","evaluation_data":null},{"case_id":"AIBD-0578","case_url":"records/AIBD-0578.html","measure":"All-cause mortality within 90 days was the primary outcome.","reported_result":"3.6% versus 4.3%","baseline":"4.3% mortality within 90 days","result":"3.6% mortality within 90 days","baseline_value":null,"result_value":null,"unit":null,"effect":"Mortality was 3.6% with alerts and 4.3% with usual care. The hazard ratio was 0.83 (95% CI, 0.70–0.99).","design":"An unblinded, pragmatic randomized trial with patients assigned to alerts or usual care. Allocation used a random list of medical-record numbers prepared before enrollment.","sample":"The analysis included 15,965 adults treated by 39 physicians at two hospitals in Taiwan.","uncertainty":"The study was unblinded and concentrated in one health system. Deaths outside its records could be missed. Survival beyond 90 days was not established.","comparison_basis":"measured","source_urls":["https://www.nature.com/articles/s41591-024-02961-4","https://www.nature.com/articles/s41746-025-01735-7","https://github.com/Imshepherd/ECGSurvNet/tree/68e3b7c58491a38d731d43626f98443e53d035a5","https://pmc.ncbi.nlm.nih.gov/articles/PMC12506480/"],"source_refs":["AIBD-0578-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Randomization supports an effect of the alert-assisted care pathway. It does not identify which subsequent clinical actions produced the survival difference.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Usual care without the AI-ECG report or high-risk alert during the 90-day follow-up.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"An unblinded, pragmatic randomized trial with patients assigned to alerts or usual care. Allocation used a random list of medical-record numbers prepared before enrollment.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mortality was 3.6% with alerts and 4.3% with usual care. The hazard ratio was 0.83 (95% CI, 0.70–0.99).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Patients entered from 15 December 2021 through 30 April 2022. A physician received an alert for the first high-risk ECG.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"NCT05118035","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Researchers compared deaths recorded in hospital records within 90 days. Patients were censored at their last known follow-up.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"All-cause mortality within 90 days was the primary outcome.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The article includes a protocol and randomization details. Training code is public. Patient data require ethics approval and secure access.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"The analysis included 15,965 adults treated by 39 physicians at two hospitals in Taiwan.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The study was unblinded and concentrated in one health system. Deaths outside its records could be missed. Survival beyond 90 days was not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0578-reported-measure","evaluation_data":null},{"case_id":"AIBD-0184","case_url":"records/AIBD-0184.html","measure":"DWP estimates that the model saved £4.4 million from 2022–23 to 2024–25","reported_result":"DWP estimates that the model saved £4.4 million from 2022–23 to 2024–25","baseline":"fraud-risk identification rate of the random control sample (not quantified)","result":"around three times more effective at identifying fraud risk","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.nao.org.uk/wp-content/uploads/2025/10/Tackling-benefit-overpayments-due-to-fraud-and-error.pdf","https://www.civilserviceworld.com/news/article/digital-and-data-challenges-could-hamper-dwps-fraud-strategy-nao-warns","https://www.gov.uk/government/publications/effectiveness-assessment-including-statistical-analysis-of-the-universal-credit-advances-machine-learning-model-1-april-2025-to-31-march-2026/effectiveness-assessment-of-universal-credit-advances-model","https://www.theguardian.com/society/2024/dec/06/revealed-bias-found-in-ai-system-used-to-detect-uk-benefits","https://www.civilserviceworld.com/news/article/what-is-the-cost-of-dwps-uc-antifraud-algorithm"],"source_refs":["AIBD-0184-source-0001","AIBD-0184-source-0002","AIBD-0184-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a randomised control group sample of advance requests","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: fraud-risk identification rate of the random control sample (not quantified). Result: around three times more effective at identifying fraud risk.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"saving an estimated £4.4 million from 2022–23 to 2024–25 by DWP's estimate","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0184-reported-measure","evaluation_data":null},{"case_id":"AIBD-0358","case_url":"records/AIBD-0358.html","measure":"received tiger notifications 30 to 42 seconds after detection","reported_result":"received tiger notifications 30 to 42 seconds after detection","baseline":null,"result":"almost wiped out false alarms","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.taipeitimes.com/News/feat/archives/2023/10/06/2003807276","https://doi.org/10.1093/biosci/biad076"],"source_refs":["AIBD-0358-source-0001","AIBD-0358-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: conventional camera traps tripped by passing boars or falling leaves","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Result: almost wiped out false alarms.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"tiger notifications arrived 30 to 42 seconds after detection","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0358-reported-measure","evaluation_data":null},{"case_id":"AIBD-0079","case_url":"records/AIBD-0079.html","measure":"thrombectomy restored flow under 60 minutes from arrival; no deficits","reported_result":"thrombectomy restored flow under 60 minutes from arrival; no deficits","baseline":"tethered to our workstations and had to take multiple steps to open the software to view the scans","result":"RapidAI alert on his smartphone and smartwatch; artery opened in under 60 minutes from arrival","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.mhs.net/news/2022/05/memorial-neuroscience-institute-uses-ai-to-optimize-care-and-positive-outcomes-for-stroke-patients"],"source_refs":["AIBD-0079-source-0001","AIBD-0079-source-0002","AIBD-0079-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the hospital's workflow before RapidAI, with scans viewed at workstations","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: tethered to our workstations and had to take multiple steps to open the software to view the scans. Result: RapidAI alert on his smartphone and smartwatch; artery opened in under 60 minutes from arrival.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"thrombectomy restored flow under 60 minutes from arrival; no deficits","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0079-reported-measure","evaluation_data":null},{"case_id":"AIBD-0601","case_url":"records/AIBD-0601.html","measure":"Pedestrian crash risk and pedestrian injury crash risk.","reported_result":"a 25% to 27% lower pedestrian crash risk and a 29% to 30% lower pedestrian injury crash risk","baseline":"Pedestrian crash risk of unequipped vehicles","result":"25% to 27% lower pedestrian crash risk","baseline_value":null,"result_value":null,"unit":null,"effect":"25% to 27% lower pedestrian crash risk; 29% to 30% lower pedestrian injury crash risk.","design":"controlled observational.","sample":"US police-reported pedestrian crashes involving vehicle models on which pedestrian-detecting AEB was optional.","uncertainty":"Observational; no evidence of benefit at night on unlit roads, at 50 mph or higher speed limits, or in turns; makers and software not named in the abstract.","comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/35580401/","https://www.iihs.org/research-areas/pedestrians-and-bicyclists"],"source_refs":["AIBD-0601-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"25% to 27% lower crash risk attributed to the system.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Same models without the system.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"25% to 27% lower pedestrian crash risk; 29% to 30% lower pedestrian injury crash risk.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Vehicles fitted with pedestrian-detecting AEB.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Pedestrian crash rates per insured vehicle year (Poisson regression) and a quasi-induced exposure comparison with a crash type that the system cannot affect.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Pedestrian crash risk and pedestrian injury crash risk.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Police crash data and insured vehicle year data; code not public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"US police-reported pedestrian crashes involving vehicle models on which pedestrian-detecting AEB was optional.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Observational; no evidence of benefit at night on unlit roads, at 50 mph or higher speed limits, or in turns; makers and software not named in the abstract.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0601-reported-measure","evaluation_data":null},{"case_id":"AIBD-0031","case_url":"records/AIBD-0031.html","measure":"Receipt of alerts, advance preparation, a post-flood welfare index, home flooding, food security, distress borrowing and physical health.","reported_result":"12.7 percentage points less likely to have their own home flooded","baseline":"control households under severe flooding, whose chance of home flooding was 23.7 percentage points higher than in a season with no flood","result":"treated households under severe flooding were 12.7 percentage points less likely to have their own home flooded","baseline_value":null,"result_value":null,"unit":null,"effect":"Post-flood welfare index up 0.3 standard deviations in flooded villages; home flooding down 12.7 percentage points under severe floods; distress borrowing down by about INR 12,600; no effect on physical health.","design":"randomized.","sample":"319 village councils (gram panchayats) in 12 districts of North Bihar, 160 treated and 159 control; panel of 5,582 households over four monsoon seasons.","uncertainty":"Working paper without peer review; outcomes are self-reported; the authors acknowledge Google.org funding; the April 2024 health findings are not repeated.","comparison_basis":"measured","source_urls":["https://www.dropbox.com/scl/fi/gpwdwa4hkm9e2e81c2ne5/floods_EWS.pdf?rlkey=3n4l804arm1zms5y0pt0ac5sp&raw=1","https://bpb-us-w2.wpmucdn.com/campuspress.yale.edu/dist/7/2986/files/2024/04/Google_Floods-bcdf49042e23059a.pdf","https://www.povertyactionlab.org/print/pdf/node/8026750"],"source_refs":["AIBD-0031-source-0003"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not measured: both arms received the same AI and CWC forecasts.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Control village councils with the same forecasts and alerts and no volunteers.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Post-flood welfare index up 0.3 standard deviations in flooded villages; home flooding down 12.7 percentage points under severe floods; distress borrowing down by about INR 12,600; no effect on physical health.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Four monsoon seasons, 2022 to 2025.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Household panel surveys at midline and endline each season, with a satellite measure of flood severity for each village and season.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Receipt of alerts, advance preparation, a post-flood welfare index, home flooding, food security, distress borrowing and physical health.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"319 village councils (gram panchayats) in 12 districts of North Bihar, 160 treated and 159 control; panel of 5,582 households over four monsoon seasons.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Working paper without peer review; outcomes are self-reported; the authors acknowledge Google.org funding; the April 2024 health findings are not repeated.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0031-reported-measure","evaluation_data":null},{"case_id":"AIBD-0114","case_url":"records/AIBD-0114.html","measure":"Time to a resistance prediction and the patient’s subsequent clinical recovery.","reported_result":"flagged ceftazidime resistance in 1 hour","baseline":"Two to three days","result":"A prediction within one hour","baseline_value":null,"result_value":null,"unit":null,"effect":"The resistance prediction was available within an hour. Clinicians changed the antibiotic, and the patient later recovered and was discharged.","design":"A single-patient clinical account without a control. The hospital also reported separate, aggregate observations about its platform.","sample":"One patient at China Medical University Hospital. The hospital separately reported more than 100,000 uses of its platform.","uncertainty":"The clinical evidence comes from the treating hospital. News reports repeat its account. The original resistance-model letter could be identified but not read in full.","comparison_basis":"assumed","source_urls":["https://health.ltn.com.tw/article/breakingnews/4102412","https://www.cmuh.cmu.edu.tw/NewsInfo/NewsArticle?no=8151","https://www.cna.com.tw/news/ahel/202210260324.aspx","https://doi.org/10.1016/j.jinf.2022.09.005","https://www.cmuh.org.tw/FileUploads/FileUpload/2022%E4%B8%AD%E5%9C%8B%E9%86%ABCSR%28ESG%29_PDF.pdf","https://pmc.ncbi.nlm.nih.gov/articles/PMC13662083/"],"source_refs":["AIBD-0114-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The evidence supports the system’s role in selecting treatment earlier. It does not establish what would have happened without that warning.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The hospital’s stated turnaround for conventional susceptibility testing. There was no controlled comparison of this patient’s recovery.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"A single-patient clinical account without a control. The hospital also reported separate, aggregate observations about its platform.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"The resistance prediction was available within an hour. Clinicians changed the antibiotic, and the patient later recovered and was discharged.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"The admission began in April 2022, pneumonia developed in late May, and discharge occurred in late July. The exact date of the AI warning is not established.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"CMUH clinical account, NewsArticle 8151","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"The hospital described the laboratory warning, antibiotic change, improvement, extubation, and discharge. Public reports do not include the full medical record.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time to a resistance prediction and the patient’s subsequent clinical recovery.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The hospital reproduces a figure from its resistance-model letter. The reviewed public materials do not provide model code, weights, or patient-level data.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One patient at China Medical University Hospital. The hospital separately reported more than 100,000 uses of its platform.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The clinical evidence comes from the treating hospital. News reports repeat its account. The original resistance-model letter could be identified but not read in full.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0114-reported-measure","evaluation_data":null},{"case_id":"AIBD-0643","case_url":"records/AIBD-0643.html","measure":"Screen-detected breast cancers.","reported_result":"the three together found 269 cancers against 250 by the radiologists alone","baseline":"250 screen-detected cancers","result":"269 screen-detected cancers with AI as a third reader","baseline_value":null,"result_value":null,"unit":null,"effect":"269 vs 250 cancers with AI as third reader (relative proportion 1.08); 261 vs 250 for one radiologist plus AI.","design":"controlled observational.","sample":"55,581 women aged 40 to 74 screened at Capio Sankt Göran Hospital, 1 April 2021 to 9 June 2022.","uncertainty":"One hospital, one AI product, one mammography vendor; interval cancers not yet reported; Lunit co-funded the study.","comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/37690911/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12818982/","https://pubmed.ncbi.nlm.nih.gov/42617468/","https://web.archive.org/web/2025/https://www.thelancet.com/journals/landig/article/PIIS2589-7500(23)00153-X/fulltext"],"source_refs":["AIBD-0643-source-0003"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"19 additional cancers.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Double reading by two radiologists.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"269 vs 250 cancers with AI as third reader (relative proportion 1.08); 261 vs 250 for one radiologist plus AI.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"AI read every examination.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Screen-detected cancers within 3 months, by reader combination.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Screen-detected breast cancers.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Registered as NCT04778670; appendix with protocol.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"55,581 women aged 40 to 74 screened at Capio Sankt Göran Hospital, 1 April 2021 to 9 June 2022.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One hospital, one AI product, one mammography vendor; interval cancers not yet reported; Lunit co-funded the study.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0643-reported-measure","evaluation_data":null},{"case_id":"AIBD-0069","case_url":"records/AIBD-0069.html","measure":"raised the number of identified people in the photo from three to six","reported_result":"raised the number of identified people in the photo from three to six","baseline":"only three people in the group photo had been identified, and Fixler was listed as unidentified","result":"the number identified doubled","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.bbc.com/news/technology-63483694","https://www.timesofisrael.com/google-engineer-identifies-anonymous-faces-in-wwii-photos-with-ai-facial-recognition/","https://www.jpost.com/business-and-innovation/tech-and-start-ups/article-710634","https://text.npr.org/1144230135","https://numberstonames.org/"],"source_refs":["AIBD-0069-source-0001","AIBD-0069-source-0002","AIBD-0069-source-0003","AIBD-0069-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0069-reported-measure","evaluation_data":null},{"case_id":"AIBD-0068","case_url":"records/AIBD-0068.html","measure":"one week earlier than routine practice and avoided whole-neuraxis radiotherapy","reported_result":"one week earlier than routine practice and avoided whole-neuraxis radiotherapy","baseline":"CGP is time-consuming, with the time for diagnosis reaching up to 4 weeks","result":"targeted treatment 1 week earlier than in the routine practice; osimertinib given 2 days after the biopsy","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10646899/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12891468/"],"source_refs":["AIBD-0068-source-0001","AIBD-0068-source-0002","AIBD-0068-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: routine practice, in which comprehensive genomic profiling precedes targeted therapy","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: CGP is time-consuming, with the time for diagnosis reaching up to 4 weeks. Result: targeted treatment 1 week earlier than in the routine practice; osimertinib given 2 days after the biopsy.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"one week earlier than routine practice and avoided whole-neuraxis radiotherapy","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0068-reported-measure","evaluation_data":null},{"case_id":"AIBD-0004","case_url":"records/AIBD-0004.html","measure":"Word error rate and words per minute.","reported_result":"produced text at 62 words per minute","baseline":"18 words per minute for a handwriting BCI","result":"62 words per minute","baseline_value":null,"result_value":null,"unit":null,"effect":"62 words per minute; word error rate 9.1% on a 50-word vocabulary and 23.8% on a 125,000-word vocabulary.","design":"case study.","sample":"One woman with bulbar-onset ALS in the BrainGate2 pilot trial at Stanford, research sessions on post-implant days 119 to 141 (late July to mid August 2022).","uncertainty":"One participant, and the authors are the developers. The team called it a proof of concept, and the sessions were supervised research sessions.","comparison_basis":"measured","source_urls":["https://europepmc.org/article/PMC/PMC10468393","https://pmc.ncbi.nlm.nih.gov/articles/PMC10468393/","https://www.bpr.org/2023-08-23/these-experimental-brain-implants-can-restore-speech-to-paralyzed-patients","https://med.stanford.edu/news/all-news/2023/08/brain-implant-speech-als.html","https://www.nature.com/articles/d41586-023-02546-0","https://pmc.ncbi.nlm.nih.gov/articles/PMC12360486/"],"source_refs":["AIBD-0004-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The previous record for any BCI, 18 words per minute with a handwriting BCI.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"62 words per minute; word error rate 9.1% on a 50-word vocabulary and 23.8% on a 125,000-word vocabulary.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Real-time decoding of attempted speech of held-out sentences, scored by word error rate and words per minute.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Word error rate and words per minute.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Neural data for 10,850 attempted sentences are public on Dryad (doi:10.5061/dryad.x69p8czpq).","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One woman with bulbar-onset ALS in the BrainGate2 pilot trial at Stanford, research sessions on post-implant days 119 to 141 (late July to mid August 2022).","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One participant, and the authors are the developers. The team called it a proof of concept, and the sessions were supervised research sessions.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0004-reported-measure","evaluation_data":null},{"case_id":"AIBD-0100","case_url":"records/AIBD-0100.html","measure":"Knowledge, birth preparation, routine and danger-sign care, newborn care, and postpartum care content.","reported_result":"improvements of 0.06–0.09 standard deviations across five measures","baseline":"Usual care, with independent enrollment in PROMPTS still possible","result":"Five indices improved by 0.06–0.09 standard deviations; danger-sign care-seeking did not differ significantly","baseline_value":null,"result_value":null,"unit":null,"effect":"Five indices improved by 0.06–0.09 standard deviations. Care-seeking for danger signs did not differ significantly.","design":"Twenty facilities introduced PROMPTS and twenty continued usual care. Assignment was stratified within counties.","sample":"There were 6,139 participants at 40 facilities. Researchers completed 5,509 of 6,128 eligible postpartum interviews.","uncertainty":"Outcomes were self-reported. The study was not powered for clinical outcomes, and a parallel nurse-training program could have influenced care.","comparison_basis":"measured","source_urls":["https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004527","https://ojs.aaai.org/index.php/AAAI/article/view/26709","https://jacarandahealth.org/ypoagriw/2023/11/PROMPTS-BROCHURE-V4.pdf","https://jacarandahealth.org/ypoagriw/2025/11/Q3-2025-Impact-Report-1.pdf","https://www.talkafrica.co.ke/ai-supported-text-service-helps-kenyan-mothers-but-information-alone-is-not-enough/"],"source_refs":["AIBD-0100-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The evaluation did not separate automated triage from information, reminders, or clinician support.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Women at control facilities were not systematically introduced to PROMPTS, although they could join independently.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Twenty facilities introduced PROMPTS and twenty continued usual care. Assignment was stratified within counties.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Five indices improved by 0.06–0.09 standard deviations. Care-seeking for danger signs did not differ significantly.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Among treatment participants surveyed during pregnancy, 1,453 of 1,700 reported joining PROMPTS.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"AIBD-0100-E01, AIBD-0100-E02","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Telephone surveys measured knowledge and reported care. Analysis compared assigned groups and accounted for clustering by facility.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Knowledge, birth preparation, routine and danger-sign care, newborn care, and postpartum care content.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The public repository contains analysis code and documentation. Participant data remain restricted under the study’s approvals.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"There were 6,139 participants at 40 facilities. Researchers completed 5,509 of 6,128 eligible postpartum interviews.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Outcomes were self-reported. The study was not powered for clinical outcomes, and a parallel nurse-training program could have influenced care.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0100-reported-measure","evaluation_data":null},{"case_id":"AIBD-0612","case_url":"records/AIBD-0612.html","measure":"No-show rate overall and by race and ethnicity.","reported_result":"the no-show rate was 33% with calls and 36% without","baseline":"36% no-show rate","result":"33% no-show rate","baseline_value":null,"result_value":null,"unit":null,"effect":"33% versus 36% overall; 36% versus 42% among Black patients.","design":"randomized.","sample":"5,840 appointments with predicted no-show risk of 15% or higher at a MetroHealth internal medicine clinic.","uncertainty":"One clinic, quality improvement design, two process changes during the initiative.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10150669/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13683762/"],"source_refs":["AIBD-0612-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"3 percentage points fewer no-shows overall.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Standard automated reminders.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"33% versus 36% overall; 36% versus 42% among Black patients.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"2,858 appointments in the augmented arm.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Scheduling records of attended and missed appointments.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"No-show rate overall and by race and ethnicity.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Supplementary material; data on request.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"5,840 appointments with predicted no-show risk of 15% or higher at a MetroHealth internal medicine clinic.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One clinic, quality improvement design, two process changes during the initiative.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0612-reported-measure","evaluation_data":null},{"case_id":"AIBD-0003","case_url":"records/AIBD-0003.html","measure":"Words per minute and word error rate.","reported_result":"decoded text at a median 78 words per minute","baseline":"about 14.2 words per minute with her assistive device","result":"a median 78.3 words per minute of decoded text, with a median word error rate of 25% over a 1,024-word vocabulary","baseline_value":null,"result_value":null,"unit":null,"effect":"Median 78.3 words per minute with a median word error rate of 25% over a 1,024-word vocabulary.","design":"case study.","sample":"One woman with anarthria and quadriplegia 18 years after a brainstem stroke, BRAVO trial (NCT03698149), sessions at her residence in 2022 and 2023.","uncertainty":"One participant, and the authors are the developers. Results come from research sessions, and the paper reports no independent use.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10826467/","https://www.yahoo.com/news/woman-paralysis-speaks-avatar-18-004307945.html","https://www.bpr.org/2023-08-23/these-experimental-brain-implants-can-restore-speech-to-paralyzed-patients","https://engineering.berkeley.edu/news/2025/03/brain-to-voice-neuroprosthesis-restores-naturalistic-speech/","https://www.nature.com/articles/d41586-023-02546-0","https://news.berkeley.edu/2025/08/13/a-stroke-stole-her-ability-to-speak-eighteen-years-later-scientists-used-ai-to-bring-it-back/"],"source_refs":["AIBD-0003-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Her usual head-tracking assistive device at 14.2 words per minute.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Median 78.3 words per minute with a median word error rate of 25% over a 1,024-word vocabulary.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Real-time decoding of silently attempted sentences into text, scored by word error rate and words per minute; synthesized speech and avatar output also assessed.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Words per minute and word error rate.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Code on GitHub (UCSF-Chang-Lab-BRAVO/multimodal-decoding); data under restricted access on request.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One woman with anarthria and quadriplegia 18 years after a brainstem stroke, BRAVO trial (NCT03698149), sessions at her residence in 2022 and 2023.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One participant, and the authors are the developers. Results come from research sessions, and the paper reports no independent use.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0003-reported-measure","evaluation_data":null},{"case_id":"AIBD-0019","case_url":"records/AIBD-0019.html","measure":"New figurative geoglyphs confirmed per year.","reported_result":"303 previously unknown figurative relief-type geoglyphs, almost doubling the known total","baseline":"18.7/y from 2004 to 2020","result":"303 new figurative geoglyphs during the 2022/23 season of field work (a 16-fold acceleration)","baseline_value":null,"result_value":null,"unit":null,"effect":"303 new figurative geoglyphs in one six-month season, which the authors call a 16-fold acceleration.","design":"before/after.","sample":"The Nazca Pampa and surrounding desert, more than 600 square kilometers; 341 candidate locations visited in the 2022/23 season.","uncertainty":"The study is the developers' own; the comparison is with earlier periods and has no concurrent control; 968 promising candidates were not yet surveyed.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11459208/","https://www.yamagata-u.ac.jp/en/information/info/20240924/","https://www.dlr.de/en/eoc/latest/news/2024/accelerated-discovery-nasca-geoglyphs-by-ai","https://archaeology.org/news/2024/09/27/more-than-300-geoglyphs-discovered-in-peru/","https://www.theartnewspaper.com/2025/08/14/248-nazca-lines-discovered-artificial-intelligence-peru","https://www.theguardian.com/world/2024/sep/26/nazca-lines-peru-new-geoglyphs","https://www.cnn.com/2024/09/27/science/ai-nazca-geoglyphs-peru"],"source_refs":["AIBD-0019-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"178 of the 303 were individually suggested by the model; 66 more were found in AI-discovered groups, and 59 were found without AI help or outside the AI-suggested boxes.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"18.7 per year from 2004 to 2020 with high-resolution imagery and no AI.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"303 new figurative geoglyphs in one six-month season, which the authors call a 16-fold acceleration.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One field season, September 2022 to February 2023, 1,440 labor hours of survey.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Archaeologists screened AI-ranked candidate boxes on 10 cm aerial images and confirmed geoglyphs by survey on foot and with drones.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"New figurative geoglyphs confirmed per year.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Data are held by the Peruvian Ministry of Culture and are available on request with its permission.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"The Nazca Pampa and surrounding desert, more than 600 square kilometers; 341 candidate locations visited in the 2022/23 season.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The study is the developers' own; the comparison is with earlier periods and has no concurrent control; 968 promising candidates were not yet surveyed.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0019-reported-measure","evaluation_data":null},{"case_id":"AIBD-0318","case_url":"records/AIBD-0318.html","measure":null,"reported_result":null,"baseline":"camera traps have lacked the real-time notification capabilities and practicality necessary for prevention of poaching","result":"notifications of tiger presence via email or push notification between 30 and 42 seconds after detections","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10580963/","https://www.taipeitimes.com/News/feat/archives/2023/10/06/2003807276","https://news.clemson.edu/clemson-students-deployment-of-novel-camera-alert-system-trailguard-ai-featured-in-bioscience-promotes-coexistence-of-tigers-humans/","https://www.thehindu.com/sci-tech/energy-and-environment/ai-cameras-to-curb-poaching/article67315766.ece"],"source_refs":["AIBD-0318-source-0001","AIBD-0318-source-0002","AIBD-0318-source-0003","AIBD-0318-source-0004","AIBD-0318-source-0005","AIBD-0318-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: camera traps without real-time notification","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: camera traps have lacked the real-time notification capabilities and practicality necessary for prevention of poaching. Result: notifications of tiger presence via email or push notification between 30 and 42 seconds after detections.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0318-reported-measure","evaluation_data":null},{"case_id":"AIBD-0022","case_url":"records/AIBD-0022.html","measure":"Apple Watch Series 7 had shown an oxygen saturation of 88% in the early morning","reported_result":"Apple Watch Series 7 had shown an oxygen saturation of 88% in the early morning","baseline":"Spot oximetry read 94-98% at primary care and 97% at the hospital, and the chest radiograph and electrocardiogram were normal","result":"The watch had recorded 88% in the early morning; CT angiography then confirmed pulmonary embolism","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC9494381/","https://europepmc.org/article/PMC/PMC12511463","https://pmc.ncbi.nlm.nih.gov/articles/PMC10039641/"],"source_refs":["AIBD-0022-source-0001","AIBD-0022-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Apple Watch Series 7 had shown an oxygen saturation of 88% in the early morning","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Apple Watch Series 7 had shown an oxygen saturation of 88% in the early morning","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0022-reported-measure","evaluation_data":null},{"case_id":"AIBD-0172","case_url":"records/AIBD-0172.html","measure":"Classification rate per species; share of inner-zone entries with an active shutdown signal.","reported_result":"Correct classification of 99.7% to 99.9% for red kite, black kite and white-tailed eagle and 97.8% for lesser spotted eagle; shutdown signal active in 90.5% of 1,054 inner-zone entries, 99.4% after 1.5 seconds.","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":"Correct classification of 99.7% to 99.9% for red kite, black kite and white-tailed eagle and 97.8% for lesser spotted eagle; shutdown signal active in 90.5% of 1,054 inner-zone entries, 99.4% after 1.5 seconds.","design":"other.","sample":"Three test sites (Arpsdorf, Bendorf, Kublank) in Schleswig-Holstein and Mecklenburg-Vorpommern in 2025; lesser spotted eagles occurred only at Kublank.","uncertainty":"Commissioned by the maker; turbines at Kublank were virtual; for lesser spotted eagle only the classification rate was tested.","comparison_basis":null,"source_urls":["http://www.oekofor.de/en/portfolio/idf_en/","https://oecd.ai/en/incidents/2022-09-20-caeb","https://www.naturschutz-energiewende.de/download/kne-uebersicht-pruef-und-validierungsberichte-fuer-antikollisionssysteme/","https://static1.squarespace.com/static/5e710b62f0a8ee01b4352bf5/t/6abd6c00b35af038ba731f82/1790798849080/20260827_Pruefung_IdentiFlightV5_KNE-Kriterien_Final.pdf","https://www.bestellen.bayern.de/med/c1e03310-1c60-11f1-88f8-d7aff5347ccf/4b0e6a70-1059-11d9-4c85-9d915831e9eb/0/lfu_nat_00501.pdf","https://pubs.usgs.gov/publication/70248352"],"source_refs":["AIBD-0172-source-0004"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not measured as fewer deaths; the study measures detection and signaling only.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Ornithologists' reference observations.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Correct classification of 99.7% to 99.9% for red kite, black kite and white-tailed eagle and 97.8% for lesser spotted eagle; shutdown signal active in 90.5% of 1,054 inner-zone entries, 99.4% after 1.5 seconds.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Test sites in 2025, including virtual turbines.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Validation against KNE criteria: system tracks compared with ornithologists' laser range finder tracks and manually checked images.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Classification rate per species; share of inner-zone entries with an active shutdown signal.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Methods and site descriptions are in the report; raw data are not published.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Three test sites (Arpsdorf, Bendorf, Kublank) in Schleswig-Holstein and Mecklenburg-Vorpommern in 2025; lesser spotted eagles occurred only at Kublank.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Commissioned by the maker; turbines at Kublank were virtual; for lesser spotted eagle only the classification rate was tested.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0172-reported-measure","evaluation_data":null},{"case_id":"AIBD-0117","case_url":"records/AIBD-0117.html","measure":null,"reported_result":null,"baseline":"Days for a radiology report, and by the authors' citation a median of about three months from radiological diagnosis to first treatment","result":"Alert during the scan and stereotactic biopsy the next day","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10550689/","https://doi.org/10.3171/CASE22536","https://www.trthaber.com/haber/saglik/tip-tarihinde-ilk-yapay-zeka-hayat-kurtardi-769140.html","https://www.haberturk.com/yapay-zekanin-erken-uyariyla-hayatini-kurtardigi-hasta-3593925","https://pmc.ncbi.nlm.nih.gov/articles/PMC12192951/"],"source_refs":["AIBD-0117-source-0002","AIBD-0117-source-0001","AIBD-0117-source-0003","AIBD-0117-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0117-reported-measure","evaluation_data":null},{"case_id":"AIBD-0311","case_url":"records/AIBD-0311.html","measure":"Daily median time from message receipt to crisis-specialist triage; sensitivity, specificity and positive predictive value of the classifier.","reported_result":"the daily median time from message receipt to crisis-specialist triage was 8 to 13 minutes","baseline":"over 9 h on average","result":"8 to 13 min","baseline_value":null,"result_value":null,"unit":null,"effect":"Daily median receipt-to-triage time of 8 to 13 minutes, against over 9 hours on average before deployment; prospective AUC 0.98 and sensitivity 0.98.","design":"before/after.","sample":"102,471 chat messages sent by 32,803 patients of Cerebral, a national United States tele-mental-health provider, from 1 to 31 October 2022.","uncertainty":"Company-authored study with no concurrent control; baseline from an unpublished internal analysis; one month of data; no patient outcome measured; prospective positive predictive value of 0.66.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10663535/","https://hai.stanford.edu/news/using-nlp-detect-mental-health-crises"],"source_refs":["AIBD-0311-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The before figure is an internal average for all chat messages, so the study does not isolate how much of the change the classifier produced.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Response times to patient messages before CMD-1, from Cerebral's internal analysis.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Daily median receipt-to-triage time of 8 to 13 minutes, against over 9 hours on average before deployment; prospective AUC 0.98 and sensitivity 0.98.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"All patient chat messages during the deployment month.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Timestamps from message receipt to the Slack post by CMD-1 and to triage by a crisis specialist, logged in production; crisis specialists labeled each surfaced message as a true or false positive.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Daily median time from message receipt to crisis-specialist triage; sensitivity, specificity and positive predictive value of the classifier.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"102,471 chat messages sent by 32,803 patients of Cerebral, a national United States tele-mental-health provider, from 1 to 31 October 2022.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Company-authored study with no concurrent control; baseline from an unpublished internal analysis; one month of data; no patient outcome measured; prospective positive predictive value of 0.66.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0311-reported-measure","evaluation_data":null},{"case_id":"AIBD-0367","case_url":"records/AIBD-0367.html","measure":"51 of the roughly 100 participants","reported_result":"51 of the roughly 100 participants","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.welt.de/regionales/nrw/article241830977/Ministerium-Palantir-klaerte-unter-anderem-Amokdrohung-auf.html","https://rp-online.de/nrw/panorama/duesseldorf-palantir-klaert-amokdrohung-und-schiesserei-in-duisburg-auf_aid-79054087","https://www.wp.de/staedte/siegerland/article236775413/Siegen-Amokdrohung-gegen-Stadtverwaltung.html","https://oecd.ai/en/incidents/2022-10-28-8649","https://freiheitsrechte.org/themen/freiheit-im-digitalen-zeitalter/stop-data-mining"],"source_refs":["AIBD-0367-source-0001","AIBD-0367-source-0002","AIBD-0367-source-0003","AIBD-0367-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"51 of the roughly 100 participants","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"51 of the roughly 100 participants","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0367-reported-measure","evaluation_data":null},{"case_id":"AIBD-0411","case_url":"records/AIBD-0411.html","measure":"Sensitivity and specificity for lesion detection and for identifying pancreatic ductal adenocarcinoma.","reported_result":"Sensitivity 34.1% and specificity 6.3% above the mean radiologist for PDAC; lesion detection sensitivity of 92.9% and specificity of 99.9% in the real-world test.","baseline":"Mean radiologist sensitivity and specificity for identifying pancreatic ductal adenocarcinoma","result":"PANDA sensitivity 34.1% higher and specificity 6.3% higher","baseline_value":null,"result_value":null,"unit":null,"effect":"Sensitivity 34.1% and specificity 6.3% above the mean radiologist for PDAC; lesion detection sensitivity of 92.9% and specificity of 99.9% in the real-world test.","design":"controlled observational.","sample":"3,208 training patients at one Shanghai center; a multicenter validation of 6,239 patients from 10 centers, including 5,337 patients from nine external centers in mainland China, Taiwan and the Czech Republic; 20,530 consecutive patients in the real-world test; 33 readers from 12 institutions in the first reader study.","uncertainty":"All tests are retrospective, the real-world test took place at one center, and eight authors are Alibaba employees. Follow-up care had already found eight of the 26 missed lesions before the study, and an independent review notes that no study shows a gain in stage, resection or survival.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10719100/","https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://pmc.ncbi.nlm.nih.gov/articles/PMC13702653/"],"source_refs":["AIBD-0411-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Five cancers and 26 other pancreatic lesions found among 20,530 patients that the initial standard of care had missed.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean performance of radiologists reading non-contrast CT; radiology reports based on contrast-enhanced CT.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sensitivity 34.1% and specificity 6.3% above the mean radiologist for PDAC; lesion detection sensitivity of 92.9% and specificity of 99.9% in the real-world test.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"PANDA applied to non-contrast abdominal or chest CT scans.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective diagnostic accuracy study with two reader studies, an external multicenter test and a retrospective real-world test against the clinical standard of truth.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sensitivity and specificity for lesion detection and for identifying pancreatic ductal adenocarcinoma.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Methods and supplementary information; the model and data are not public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"3,208 training patients at one Shanghai center; a multicenter validation of 6,239 patients from 10 centers, including 5,337 patients from nine external centers in mainland China, Taiwan and the Czech Republic; 20,530 consecutive patients in the real-world test; 33 readers from 12 institutions in the first reader study.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"All tests are retrospective, the real-world test took place at one center, and eight authors are Alibaba employees. Follow-up care had already found eight of the 26 missed lesions before the study, and an independent review notes that no study shows a gain in stage, resection or survival.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0411-reported-measure","evaluation_data":null},{"case_id":"AIBD-0645","case_url":"records/AIBD-0645.html","measure":"Adenoma detection rate.","reported_result":"adenoma detection rose from 50.7% to 54.9%","baseline":"Adenoma detection rate of 50.7% at equipped facilities before deployment","result":"Adenoma detection rate of 54.9% after deployment, against 51.1% at control facilities","baseline_value":null,"result_value":null,"unit":null,"effect":"+4.2 percentage points (50.7% to 54.9%) at equipped facilities against -0.7 points at control facilities; odds ratio 1.22.","design":"randomized.","sample":"269 endoscopists at 42 equipped facilities and 547 at 97 control facilities; 334,200 colonoscopies.","uncertainty":"Facility-level randomization; no data on cancer outcomes; deskilling not assessed.","comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/42250891/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13604928/"],"source_refs":["AIBD-0645-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Odds ratio 1.22 (95% CI 1.15 to 1.28).","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Control facilities with a random deployment date.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"+4.2 percentage points (50.7% to 54.9%) at equipped facilities against -0.7 points at control facilities; odds ratio 1.22.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Device available for all colonoscopies at equipped facilities.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Adenoma detection from VA colonoscopy and pathology records before and after deployment; mixed effects logistic regression.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Adenoma detection rate.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Registered as NCT05888623.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"269 endoscopists at 42 equipped facilities and 547 at 97 control facilities; 334,200 colonoscopies.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Facility-level randomization; no data on cancer outcomes; deskilling not assessed.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0645-reported-measure","evaluation_data":null},{"case_id":"AIBD-0020","case_url":"records/AIBD-0020.html","measure":"multiplying two 4x4 matrices over the field with two elements in 47 multiplications","reported_result":"multiplying two 4x4 matrices over the field with two elements in 47 multiplications","baseline":"49 multiplications","result":"47 multiplications","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://arxiv.org/pdf/2212.01175","https://www.nature.com/articles/s41586-022-05172-4","https://arxiv.org/abs/2210.04045"],"source_refs":["AIBD-0020-source-0001","AIBD-0020-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"multiplying two 4x4 matrices over the field with two elements in 47 multiplications","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"multiplying two 4x4 matrices over the field with two elements in 47 multiplications","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0020-reported-measure","evaluation_data":null},{"case_id":"AIBD-0228","case_url":"records/AIBD-0228.html","measure":"2,780,756 registrations (22.6%) were identified as ineligible and invalidated","reported_result":"2,780,756 registrations (22.6%) were identified as ineligible and invalidated","baseline":"12,298,944 Nigerians successfully completed the registration as new voters","result":"the number of valid registrations (post-ABIS) is 9,518,188","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.ndr.org.ng/nigerias-voter-register-hits-93-5-million-as-12m-new-registrants-are-added/","https://punchng.com/double-registration-inec-delists-1-1m-newly-registered-voters/","https://businessday.ng/uncategorized/article/eight-questions-inec-must-answer-before-the-first-ballot-is-cast/","https://web.archive.org/web/20261004011619id_/https://www.eeas.europa.eu/sites/default/files/documents/2023/EU%20EOM%20NGA%202023%20FR.pdf"],"source_refs":["AIBD-0228-source-0001","AIBD-0228-source-0002","AIBD-0228-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the raw Continuous Voter Registration data before ABIS cleaning","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 12,298,944 Nigerians successfully completed the registration as new voters. Result: the number of valid registrations (post-ABIS) is 9,518,188.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"2,780,756 registrations (22.6%) were identified as ineligible and invalidated","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0228-reported-measure","evaluation_data":null},{"case_id":"AIBD-0098","case_url":"records/AIBD-0098.html","measure":"installed shelter kits for 78 Aurora households before Storm Nalgae","reported_result":"installed shelter kits for 78 Aurora households before Storm Nalgae","baseline":"100 pre-identified households planned to receive shelter strengthening kits, with a 72-hour lead time","result":"78 households reached in two barangays of Dipaculao, with a 56-hour lead time","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://go-api.ifrc.org/api/DownloadFile/67581/MDRPH049eapfr","https://adore.ifrc.org/Download.aspx?FileId=590819","https://nhess.copernicus.org/articles/24/309/2024/","https://meetingorganizer.copernicus.org/EGU22/EGU22-10474.html"],"source_refs":["AIBD-0098-source-0001","AIBD-0098-source-0002","AIBD-0098-source-0003","AIBD-0098-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"installed shelter kits for 78 Aurora households before Storm Nalgae","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"installed shelter kits for 78 Aurora households before Storm Nalgae","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0098-reported-measure","evaluation_data":null},{"case_id":"AIBD-0059","case_url":"records/AIBD-0059.html","measure":"Word-level recognition accuracy.","reported_result":"identified 680 of 936 test utterances (72.65%)","baseline":"mean = 56.75 %, SD = 12.91","result":"72.65 % accuracy","baseline_value":null,"result_value":null,"unit":null,"effect":"72.65% for the recognizer against a listener mean of 56.75%.","design":"case study.","sample":"One 34-year-old woman with global aphasia and dysarthria in Messina, Italy; 936 test utterances of 13 words; 12 professional listeners with 26 recordings each.","uncertainty":"One participant and a 13-word vocabulary. The listeners heard different items in a different format, and the authors state that the items were not logged for a paired comparison.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC13271983/","https://doi.org/10.2196/preprints.81999"],"source_refs":["AIBD-0059-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean accuracy of 12 familiar rehabilitation professionals on a separate listening task.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"72.65% for the recognizer against a listener mean of 56.75%.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Word-level accuracy of a speaker-dependent recognizer on held-out recordings, set against open-response transcription by familiar rehabilitation professionals.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Word-level recognition accuracy.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One 34-year-old woman with global aphasia and dysarthria in Messina, Italy; 936 test utterances of 13 words; 12 professional listeners with 26 recordings each.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One participant and a 13-word vocabulary. The listeners heard different items in a different format, and the authors state that the items were not logged for a paired comparison.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0059-reported-measure","evaluation_data":null},{"case_id":"AIBD-0247","case_url":"records/AIBD-0247.html","measure":null,"reported_result":null,"baseline":"People left inside a crossing are noticed by the driver's sight or the emergency button","result":"The camera system lights the special signal lamp after a person or bicycle stays three seconds inside the closed crossing","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.yomiuri.co.jp/national/20260919-GYT1T00229/2/","https://response.jp/article/2022/11/11/363968.html","https://oecd.ai/en/incidents/2022-11-09-4a03","https://www.asahi.com/articles/ASQCH6CY1QCGUTIL005.html"],"source_refs":["AIBD-0247-source-0001","AIBD-0247-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0247-reported-measure","evaluation_data":null},{"case_id":"AIBD-0259","case_url":"records/AIBD-0259.html","measure":"Risk of a rear-end-striking crash of any severity, with any injuries, and with serious or fatal injuries.","reported_result":"front-to-rear crashes were reduced by 49%","baseline":"not stated","result":"front to rear crashes were reduced by 49 percent","baseline_value":null,"result_value":null,"unit":null,"effect":"AEB: 34% lower risk of a rear-end-striking crash of any severity and with any injuries; 76% lower risk with serious or fatal injuries (p = 0.09). FCW: 22% lower risk with any injuries and 71% with serious or fatal injuries; no change in all rear-end-striking crashes.","design":"controlled observational.","sample":"Pickups with and without optional FCW or AEB in 25 U.S. states, 2017 to 2020.","uncertainty":"Observational; the serious-injury estimate for AEB was not statistically significant (p = 0.09); the systems were optional equipment on the pickups studied.","comparison_basis":"measured","source_urls":["https://learningenglish.voanews.com/a/study-vehicle-warning-and-braking-systems-cut-crashes-by-50-percent/6836920.html","https://oecd.ai/en/incidents/2022-11-15-d32d","https://www.safetyresearch.net/wp-content/uploads/2025/04/PARTS_Real-World-Effectiveness-ADAS.pdf","https://pubmed.ncbi.nlm.nih.gov/36853168/","https://www.federalregister.gov/documents/2024/05/09/2024-09054/federal-motor-vehicle-safety-standards-automatic-emergency-braking-systems-for-light-vehicles"],"source_refs":["AIBD-0259-source-0004"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Pickups of the same kind without the systems.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"AEB: 34% lower risk of a rear-end-striking crash of any severity and with any injuries; 76% lower risk with serious or fatal injuries (p = 0.09). FCW: 22% lower risk with any injuries and 71% with serious or fatal injuries; no change in all rear-end-striking crashes.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Police-reported crashes in 25 states, 2017 to 2020.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Police-reported rear-end crashes with a pickup as the striking vehicle, linked to optional FCW or AEB equipment; quasi-induced exposure with logistic regression and crash rates per registered vehicle year with Poisson regression.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Risk of a rear-end-striking crash of any severity, with any injuries, and with serious or fatal injuries.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Pickups with and without optional FCW or AEB in 25 U.S. states, 2017 to 2020.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Observational; the serious-injury estimate for AEB was not statistically significant (p = 0.09); the systems were optional equipment on the pickups studied.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0259-reported-measure","evaluation_data":null},{"case_id":"AIBD-0332","case_url":"records/AIBD-0332.html","measure":"the audit achieved corrections of notices worth R$91,723,011.31","reported_result":"the audit achieved corrections of notices worth R$91,723,011.31","baseline":"60 irregularities involving R$113,713,128.26","result":"53 alerts answered and notices worth R$91,723,011.31 corrected or addressed in later tenders","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://atricon.org.br/blog/tce-mg-apresenta-economia-feita-com-ajuda-de-robo-alice/","https://www.tce.mg.gov.br/Tribunal-mineiro-apresenta-economia-feita-com-ajuda-de-Alice.html/Noticia/1111626457","https://atricon.org.br/blog/tce-mg-apresenta-robos-alice-e-solaris-para-o-tce-es/","https://ofator.com.br/informacao/como-funcionam-as-ias-desenvolvidas-pelo-tce-para-fiscalizar-licitacoes-em-minas-gerais/"],"source_refs":["AIBD-0332-source-0001","AIBD-0332-source-0002","AIBD-0332-source-0003","AIBD-0332-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the audit achieved corrections of notices worth R$91,723,011.31","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the audit achieved corrections of notices worth R$91,723,011.31","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0332-reported-measure","evaluation_data":null},{"case_id":"AIBD-0067","case_url":"records/AIBD-0067.html","measure":"Completion of a diabetic eye exam within six months, with referral attendance as a secondary outcome.","reported_result":"all 81 young people offered IDx-DR completed an eye exam, compared with 18 of 82 controls","baseline":"22%, 95%CI: 14.2%, 32.4%","result":"100%, 95%CI: 95.5%, 100%","baseline_value":null,"result_value":null,"unit":null,"effect":"Screening completion was 100% with AI and 22% with referral, a difference of 78 percentage points (95% CI, 69–87).","design":"An unmasked randomized trial, with allocation concealed until assignment. Randomization used blocks of four or six and was stratified by clinic.","sample":"The trial randomized 164 participants at two Baltimore clinics. After one control was excluded, the analysis included 81 AI participants and 82 controls.","uncertainty":"The trial measured access to screening, not later vision. The referral comparison used 25 AI-positive participants versus the entire control group, so those secondary denominators are not equivalent.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10784572/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13272687/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11825398/"],"source_refs":["AIBD-0067-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Randomization supports the effect of offering an immediate AI exam rather than a referral. It does not separate the algorithm from convenience and access at the clinic.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Scripted education and referral to an eye-care provider, with a printed handout.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"An unmasked randomized trial, with allocation concealed until assignment. Randomization used blocks of four or six and was stratified by clinic.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Screening completion was 100% with AI and 22% with referral, a difference of 78 percentage points (95% CI, 69–87).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Participants enrolled from 24 November 2021 to 6 June 2022. Follow-up ended on 6 December 2022. An AI exam generally took five to ten minutes.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"NCT05131451","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"The primary outcome was documented completion of a diabetic eye exam within six months. Researchers also tracked referral attendance and participant satisfaction.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Completion of a diabetic eye exam within six months, with referral attendance as a secondary outcome.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The protocol and supplement describe the procedure. De-identified data require an approved proposal and data-use agreement. The commercial diagnostic code is not published.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"The trial randomized 164 participants at two Baltimore clinics. After one control was excluded, the analysis included 81 AI participants and 82 controls.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The trial measured access to screening, not later vision. The referral comparison used 25 AI-positive participants versus the entire control group, so those secondary denominators are not equivalent.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0067-reported-measure","evaluation_data":null},{"case_id":"AIBD-0579","case_url":"records/AIBD-0579.html","measure":"In-hospital mortality, completion of the sepsis care bundle, and change in organ dysfunction over 72 hours.","reported_result":"1.9 percentage points","baseline":"11.39% modeled counterfactual mortality","result":"9.49% observed post-deployment mortality","baseline_value":null,"result_value":null,"unit":null,"effect":"Mortality was 9.49%, versus 11.39% in the modeled comparison: a difference of 1.9 percentage points (95% CI, 0.3–3.5).","design":"A prospective, nonrandomized before-and-after study. A Bayesian time-series model estimated what would have happened without deployment.","sample":"The analysis included 6,217 eligible sepsis encounters: 5,065 before deployment and 1,152 afterward.","uncertainty":"There was no randomized or concurrent untreated group. Unmeasured changes in care could affect the comparison. Results covered five months at one health system.","comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10805720/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11169238/","https://pmc.ncbi.nlm.nih.gov/articles/PMC8429719/","https://pmc.ncbi.nlm.nih.gov/articles/PMC10907626/","https://pubmed.ncbi.nlm.nih.gov/40360868/"],"source_refs":["AIBD-0579-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The estimate concerns COMPOSER, nurse alerts, preparation, and subsequent care together. It does not isolate the prediction model from those changes.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"A modeled post-deployment counterfactual based on the historical period, with adjustment for measured differences.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"A prospective, nonrandomized before-and-after study. A Bayesian time-series model estimated what would have happened without deployment.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mortality was 9.49%, versus 11.39% in the modeled comparison: a difference of 1.9 percentage points (95% CI, 0.3–3.5).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"The historical period ran from 1 January 2021 to 6 December 2022. Deployment began on 7 December 2022.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"UCSD IRB 805726","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Researchers used electronic records to identify Sepsis-3 cases, inpatient deaths, organ dysfunction, and completion of the sepsis care bundle.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"In-hospital mortality, completion of the sepsis care bundle, and change in organ dysfunction over 72 hours.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The supplement describes implementation and statistical methods. De-identified data and analysis code are available only through the authors and institutional approval.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"The analysis included 6,217 eligible sepsis encounters: 5,065 before deployment and 1,152 afterward.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"There was no randomized or concurrent untreated group. Unmeasured changes in care could affect the comparison. Results covered five months at one health system.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0579-reported-measure","evaluation_data":null},{"case_id":"AIBD-0613","case_url":"records/AIBD-0613.html","measure":"Screen-reading workload; cancer detection rate; interval cancer rate.","reported_result":"a 44.2% reduction in workload","baseline":"109,692 screen readings","result":"61,248 screen readings","baseline_value":null,"result_value":null,"unit":null,"effect":"44.2% fewer readings; detection 6.4 versus 5.0 per 1,000 (ratio 1.29).","design":"randomized.","sample":"105,934 women at four screening sites in southwest Sweden.","uncertainty":"One region of Sweden, one vendor version; author interests not read in full text.","comparison_basis":"measured","source_urls":["https://europepmc.org/abstract/MED/39904652","https://pmc.ncbi.nlm.nih.gov/articles/PMC13605948/","https://europepmc.org/abstract/MED/41620232"],"source_refs":["AIBD-0613-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"48,444 fewer screen readings.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Standard double reading without AI.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"44.2% fewer readings; detection 6.4 versus 5.0 per 1,000 (ratio 1.29).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"53,043 women in the AI arm.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Count of screen readings; cancer registry and pathology follow-up for detected and interval cancers.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Screen-reading workload; cancer detection rate; interval cancer rate.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"ClinicalTrials.gov NCT04838756.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"105,934 women at four screening sites in southwest Sweden.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One region of Sweden, one vendor version; author interests not read in full text.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0613-reported-measure","evaluation_data":null},{"case_id":"AIBD-0642","case_url":"records/AIBD-0642.html","measure":"Study 1: Interval cancer rate per 1,000 participants. Study 2: Cancer detection rate per 1,000 and screen-reading workload.","reported_result":"a detection rate of 6.4 per 1,000 against 5.0 per 1,000","baseline":"5.0 screen-detected cancers per 1,000 women","result":"6.4 screen-detected cancers per 1,000 women","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: 1.55 vs 1.76 per 1,000; proportion ratio 0.88 (95% CI 0.65 to 1.18), non-inferior; sensitivity 80.5% vs 73.8%. Study 2: 6.4 vs 5.0 per 1,000 (ratio 1.29); 44.2% fewer screen readings.","design":"Study 1: randomized. Study 2: randomized.","sample":"Study 1: 105,934 women screened at four sites in southwest Sweden, 12 April 2021 to 7 December 2022. Study 2: 105,915 women analyzed.","uncertainty":"Study 1: Interval cancer difference not statistically significant; one AI product; mortality not measured. Study 2: Some extra cancers may be overdiagnosed in situ disease.","comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/39904652/","https://pubmed.ncbi.nlm.nih.gov/41620232/","https://pubmed.ncbi.nlm.nih.gov/37541274/","https://pubmed.ncbi.nlm.nih.gov/42617468/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13605948/"],"source_refs":["AIBD-0642-source-0001","AIBD-0642-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Sensitivity 6.7 percentage points higher with AI. Study 2: 1.4 more cancers per 1,000 women.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Standard double reading without AI. Study 2: Standard double reading without AI.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: randomized. Study 2: randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 1.55 vs 1.76 per 1,000; proportion ratio 0.88 (95% CI 0.65 to 1.18), non-inferior; sensitivity 80.5% vs 73.8%. Study 2: 6.4 vs 5.0 per 1,000 (ratio 1.29); 44.2% fewer screen readings.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: AI triage plus detection support for all examinations in the intervention group. Study 2: AI triage plus detection support.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: randomized. Study 2: randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Interval cancers from registry follow-up; sensitivity and specificity. Study 2: Screen-detected cancers, recalls and screen readings.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Interval cancer rate per 1,000 participants. Study 2: Cancer detection rate per 1,000 and screen-reading workload.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Trial registered as NCT04838756. Study 2: Trial registered as NCT04838756.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 105,934 women screened at four sites in southwest Sweden, 12 April 2021 to 7 December 2022. Study 2: 105,915 women analyzed.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Interval cancer difference not statistically significant; one AI product; mortality not measured. Study 2: Some extra cancers may be overdiagnosed in situ disease.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0642-reported-measure","evaluation_data":null},{"case_id":"AIBD-0264","case_url":"records/AIBD-0264.html","measure":"Median phoneme scores at 40 to 85 dB SPL and in noise; share of patients keeping each map.","reported_result":"100% of the manual group kept the FOX map and 82% performed better with it","baseline":"manual arm at 3 months: median phoneme scores of 45%, 50% and 40% at 55, 70 and 85 dB SPL","result":"FOX arm at 3 months: median phoneme scores of 50%, 70% and 50% at the same levels","baseline_value":null,"result_value":null,"unit":null,"effect":"Better speech intelligibility in the FOX arm at 3 and 12 months (p < 0.05); after crossover all manual-arm patients kept the FOX map and 82% performed better with it, while 63% of FOX-arm patients kept the manual map for comfort.","design":"randomized.","sample":"24 patients receiving a first cochlear implant at Cliniques universitaires Saint-Luc, Brussels, randomized to FOX or manual fitting for one year, then switched.","uncertainty":"Small single-center sample with no masking listed in the registry; comfort favored manual maps in the FOX arm; a co-author is from the Eargroup, whose results Otoconsult presents as FOX's clinical evidence.","comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/36607743/","https://doi.org/10.1097/aud.0000000000001314","https://doi.org/10.1097/mao.0000000000003810","https://doi.org/10.1080/14670100.2019.1667574","https://web.archive.org/web/20251108111425/https://otoconsult.com/products/fox/","https://pubmed.ncbi.nlm.nih.gov/33027199/"],"source_refs":["https://pubmed.ncbi.nlm.nih.gov/36607743/"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The randomized comparison attributes the difference in measured hearing to the FOX maps.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Manual fitting by audiologists.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Better speech intelligibility in the FOX arm at 3 and 12 months (p < 0.05); after crossover all manual-arm patients kept the FOX map and 82% performed better with it, while 63% of FOX-arm patients kept the manual map for comfort.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One year of fitting in each arm.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Pure-tone thresholds, loudness scaling, spectral discrimination, word recognition in quiet and noise and speech tracking at 3 and 12 months; Mann-Whitney tests with Holm correction; map choice after crossover.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Median phoneme scores at 40 to 85 dB SPL and in noise; share of patients keeping each map.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"24 patients receiving a first cochlear implant at Cliniques universitaires Saint-Luc, Brussels, randomized to FOX or manual fitting for one year, then switched.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small single-center sample with no masking listed in the registry; comfort favored manual maps in the FOX arm; a co-author is from the Eargroup, whose results Otoconsult presents as FOX's clinical evidence.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0264-reported-measure","evaluation_data":null},{"case_id":"AIBD-0115","case_url":"records/AIBD-0115.html","measure":"a 99 percent stenosis of the left anterior descending artery","reported_result":"a 99 percent stenosis of the left anterior descending artery","baseline":"moderate coronary risk with about 50 percent narrowing, follow-up only and no invasive test","result":"a 72 percent AI probability of severe narrowing, then angiography that found a 99 percent stenosis of the left anterior descending artery","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/2026/https://www.cmuh.cmu.edu.tw/NewsInfo/NewsArticle?no=8228","https://www.cmuh.cmu.edu.tw/NewsInfo/NewsArticle?no=8228","https://health.ltn.com.tw/article/breakingnews/4168516","https://tw.news.yahoo.com/75%E6%AD%B2%E7%94%B7%E8%B5%B7%E5%BA%8A%E8%83%B8%E6%82%B6%E5%BF%83%E6%82%B8-%E4%B8%AD%E5%9C%8B%E9%99%84%E9%86%AB%E9%81%8B%E5%8B%95%E5%BF%83%E9%9B%BB%E5%9C%96ai%E8%BC%94%E5%8A%A9%E5%88%A4%E8%AE%80%E7%B3%BB%E7%B5%B1%E5%8F%8A%E6%99%82%E6%95%91%E5%91%BD-083734634.html","https://tw.news.yahoo.com/%E7%94%B7%E5%AD%90%E4%BB%A5%E7%82%BA%E8%83%83%E9%A3%9F%E9%81%93%E9%80%86%E6%B5%81-ai%E5%8F%8A%E6%99%82%E5%88%A4%E6%96%B7-%E8%A1%80%E7%AE%A1%E5%A0%B5%E5%A1%9E72-%E9%80%83%E9%81%8E%E7%8C%9D%E6%AD%BB-083151068.html","https://doi.org/10.3389/frai.2025.1496109"],"source_refs":["AIBD-0115-source-0002","AIBD-0115-source-0003","AIBD-0115-source-0004","AIBD-0115-source-0001","AIBD-0115-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"angiography found a 99% LAD stenosis, stented","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"angiography found a 99% LAD stenosis, stented","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0115-reported-measure","evaluation_data":null},{"case_id":"AIBD-0039","case_url":"records/AIBD-0039.html","measure":"Post-test scores, time spent learning, and students’ ratings of engagement, motivation, enjoyment, and growth mindset.","reported_result":"median learning gains over double those of the class session","baseline":"M = 3.5, N = 174","result":"M = 4.5, N = 142","baseline_value":null,"result_value":null,"unit":null,"effect":"Median post-test scores were 4.5 with AI and 3.5 in class. The adjusted advantage was 0.63 standard deviations.","design":"A randomized crossover trial across two lessons. Peer groups of two or three students were assigned together, then switched conditions the following week.","sample":"Harvard’s introductory physics course for life-science students. The paper reports 194 eligible students and compares 142 AI post-tests with 174 classroom post-tests.","uncertainty":"The study covered two lessons at one university. High scores limited the test’s ability to distinguish further gains.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12179260/","https://hechingerreport.org/proof-points-ai-tutor-harvard-physics/","https://news.harvard.edu/gazette/story/2024/09/professor-tailored-ai-tutor-to-physics-course-engagement-doubled/"],"source_refs":["AIBD-0039-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The AI lesson produced higher scores than the classroom lesson. The comparison does not separate the model’s contribution from the videos, prepared solutions, or self-paced format.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"An active-learning class covered the same material with peer discussion and instructor feedback. The pooled pre-test median was 2.75 points.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"A randomized crossover trial across two lessons. Peer groups of two or three students were assigned together, then switched conditions the following week.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Median post-test scores were 4.5 with AI and 3.5 in class. The adjusted advantage was 0.63 standard deviations.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Each participant was assigned one AI lesson and one classroom lesson in consecutive weeks. The AI lessons took a median of 49 minutes.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Harvard PS2 tutoring study, protocol IRB23-0797.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Researchers used quizzes before and after each lesson, logged time with the tutor, and asked students about their experience on a five-point scale.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Post-test scores, time spent learning, and students’ ratings of engagement, motivation, enjoyment, and growth mindset.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The supplement contains teaching prompts and methods. The authors also published a Stata data file on GitHub.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Harvard’s introductory physics course for life-science students. The paper reports 194 eligible students and compares 142 AI post-tests with 174 classroom post-tests.","editor_note":"Methods reports 194 eligible students. Results gives 142 AI post-tests and 174 classroom post-tests. Whether these counts describe the same analysis population is unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The study covered two lessons at one university. High scores limited the test’s ability to distinguish further gains.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0039-reported-measure","evaluation_data":{"learning":{"measure":"Post-test score compared with the pooled pre-test baseline","unit":"points","AI_post_test_median":4.5,"classroom_post_test_median":3.5,"pooled_pre_test_median":2.75,"pooled_pre_test_observations":316},"time":{"unit":"minutes","AI_median":49,"classroom_assumed_learning_time":60,"AI_proportion_under_60_minutes":0.7},"experience":{"scale":"Five-point agreement scale","engagement_mean":{"AI":4.1,"classroom":3.6},"motivation_mean":{"AI":3.4,"classroom":3.1},"enjoyment_and_growth_mindset":"No statistically significant difference."}}},{"case_id":"AIBD-0060","case_url":"records/AIBD-0060.html","measure":null,"reported_result":null,"baseline":"But Alex still had no diagnosis that explained all his symptoms.","result":"when ChatGPT suggested a diagnosis of tethered cord syndrome, 'it made a lot of sense,'","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.today.com/health/mom-chatgpt-diagnosis-pain-rcna101843","https://mcdn.podbean.com/mf/web/734it356gemnyjkf/Episode_24_Gilmer_and_Hofmann.pdf"],"source_refs":["AIBD-0060-source-0001","AIBD-0060-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: 17 physicians consulted over three years","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: But Alex still had no diagnosis that explained all his symptoms. Result: when ChatGPT suggested a diagnosis of tethered cord syndrome, 'it made a lot of sense,'.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0060-reported-measure","evaluation_data":null},{"case_id":"AIBD-0335","case_url":"records/AIBD-0335.html","measure":"Number of pools taxed under property tax; recurring revenue and back-tax collected.","reported_result":"140,000 taxable pools in 2023","baseline":"le géomètre qui remarquait une nouvelle piscine dans un quartier et devait vérifier dans le cadastre si elle était déclarée ou pas","result":"Dans toute la France, 140.000 piscines clandestines ont été repérées, ce qui va rapporter 40 millions d'euros.","baseline_value":null,"result_value":null,"unit":null,"effect":"+20,356 taxed pools (+5.7%); €4.4 million recurring revenue and €5.7 million in back-tax in 2022.","design":"before/after.","sample":"Nine pilot départements, 2022; national rollout in 2023 covered only by the DGFiP estimate.","uncertainty":"The Cour reports that the tool produces many false positives and has a larger error margin than detection by cadastral surveyors. It notes that the DITP received no figures on staff savings for the tool and that the DGFiP never explained its first target of €130 million a year.","comparison_basis":"assumed","source_urls":["https://www.ici.fr/infos/societe/le-fisc-decouvre-1500-piscines-non-declarees-en-dordogne-grace-a-l-intelligence-artificielle-4899279","https://www.impots.gouv.fr/actualite/generalisation-du-foncier-innovant","https://rmc.bfmtv.com/actualites/economie/grace-a-l-ia-l-etat-decouvre-140-000-piscines-non-declarees-et-recupere-40-millions-d-euros_AV-202403210525.html","https://www.ouest-france.fr/economie/budget/garages-verandas-piscines-comment-le-fisc-repere-les-batiments-non-declares-46394b0a-e780-11ee-95e5-de9a0852a727","https://www.cnews.fr/france/2024-06-11/piscines-non-declarees-le-fisc-repere-140000-fraudeurs-grace-lia-1511264","https://www.20minutes.fr/economie/4095936-20240612-taxe-fonciere-grace-ia-fisc-repere-140-000-piscines-non-declarees-2023","https://www.economie.gouv.fr/files/files/directions_services/dgfip/Rapport/2023/ra_2023.pdf","https://www.ccomptes.fr/sites/default/files/2024-10/20241022-S2024-1165-L-intelligence-artificielle-dans-les-poltiques-publiques-exemple-du-MEF.pdf","https://www.franceinfo.fr/economie/impots/pergolas-verandas-le-fisc-cible-desormais-les-extensions-non-declarees-a-l-aide-d-images-satellites_7937237.html"],"source_refs":["AIBD-0335-source-0009"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Pools taxed in the same départements before the tool.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"+20,356 taxed pools (+5.7%); €4.4 million recurring revenue and €5.7 million in back-tax in 2022.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"The Cour des comptes reports figures that the DGFiP sent on pools taxed and revenue collected in the pilot départements.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Number of pools taxed under property tax; recurring revenue and back-tax collected.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Nine pilot départements, 2022; national rollout in 2023 covered only by the DGFiP estimate.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The Cour reports that the tool produces many false positives and has a larger error margin than detection by cadastral surveyors. It notes that the DITP received no figures on staff savings for the tool and that the DGFiP never explained its first target of €130 million a year.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0335-reported-measure","evaluation_data":null},{"case_id":"AIBD-0336","case_url":"records/AIBD-0336.html","measure":"135 cases with revenue (benefits stopped or reduced)","reported_result":"135 cases with revenue (benefits stopped or reduced)","baseline":null,"result":"33% of requested cases ended with revenue for Really Single, 7% for Model Abroad and 15% for the parental leave controls","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.amnesty.org/en/wp-content/uploads/2024/11/EUR1887092024ENGLISH.pdf","https://www.amnesty.org/en/documents/eur18/8709/2024/en/","https://www.amnesty.org/en/latest/news/2024/11/denmark-ai-powered-welfare-system-fuels-mass-surveillance-and-risks-discriminating-against-marginalized-groups/","https://offentlig-ai.dk/projekter/kontrol-med-udbetaling-ydelser","https://algorithmwatch.org/en/udbetaling-danmark/","https://automatingsociety.algorithmwatch.org/report2020/denmark/","https://gijn.org/stories/amnesty-internationals-investigation-algorithms-denmarks-welfare-system/","http://justitia-int.org/wp-content/uploads/2019/07/Analyse-Udbetaling-Danmark-systematiske-overva%CC%8Agning.pdf"],"source_refs":["AIBD-0336-source-0002","AIBD-0336-source-0001","AIBD-0336-source-0003","AIBD-0336-source-0004","AIBD-0336-source-0005","AIBD-0336-source-0006","AIBD-0336-source-0007","AIBD-0336-source-0008","AIBD-0336-source-0009"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"2023 figures of 135, 36 and 72 cases with revenue from three models","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"2023 figures of 135, 36 and 72 cases with revenue from three models","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0336-reported-measure","evaluation_data":null},{"case_id":"AIBD-0472","case_url":"records/AIBD-0472.html","measure":null,"reported_result":null,"baseline":"Inspection was labor-intensive and invasive and gave little real-time data, according to the developer.","result":"The robot inspected more than 30 km of mains in four towns and gathered image and sound data, according to the developer.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.aihub.or.kr/aihubdata/data/view.do?currMenu=115&topMenu=100&dataSetSn=71737"],"source_refs":["AIBD-0472-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0472-reported-measure","evaluation_data":null},{"case_id":"AIBD-0522","case_url":"records/AIBD-0522.html","measure":"Weighted total score out of 5.","reported_result":"vivo ranked first with 3.513, which is 1.394 points above the baseline; the expert group scored cultural adaptability lowest at 2.36.","baseline":"The baseline system from Beijing Tian Tang Technology scored 2.119 of 5.","result":"The vivo system scored 3.513 of 5.","baseline_value":null,"result_value":null,"unit":null,"effect":"vivo ranked first with 3.513, which is 1.394 points above the baseline; the expert group scored cultural adaptability lowest at 2.36.","design":"other.","sample":"Ten teams submitted systems and four passed preliminary screening; 18 professional evaluators from the Chinese Association of the Deaf's Committee for Sign Language Research and Promotion and 9 general evaluators rated the output; the training set held 1,074 text and video sentence pairs.","uncertainty":"The organizers report no confidence intervals. The test set held four sentences in an evaluation setting, and it does not measure whether deaf users communicate better in daily life.","comparison_basis":"measured","source_urls":["https://aclanthology.org/2024.ccl-3.45.pdf","https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.vivo.com.cn/brand/news/detail?id=1245"],"source_refs":["AIBD-0522-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The baseline system from Beijing Tian Tang Technology (2.119) and the other submitted systems.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"vivo ranked first with 3.513, which is 1.394 points above the baseline; the expert group scored cultural adaptability lowest at 2.36.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"A shared-task evaluation in which panels of sign language users rated each system's sign language avatar output on test sentences for grammar accuracy, naturalness, readability and cultural adaptability.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Weighted total score out of 5.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Ten teams submitted systems and four passed preliminary screening; 18 professional evaluators from the Chinese Association of the Deaf's Committee for Sign Language Research and Promotion and 9 general evaluators rated the output; the training set held 1,074 text and video sentence pairs.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The organizers report no confidence intervals. The test set held four sentences in an evaluation setting, and it does not measure whether deaf users communicate better in daily life.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0522-reported-measure","evaluation_data":null},{"case_id":"AIBD-0617","case_url":"records/AIBD-0617.html","measure":"Probability of earning an A or B.","reported_result":"raised the likelihood that students earned an A or B by four percentage points","baseline":"61 percent earned an A or B (control group)","result":"Four percentage points more likely to earn an A or B","baseline_value":null,"result_value":null,"unit":null,"effect":"Four percentage points from a control rate of 61 percent.","design":"randomized.","sample":"1,568 students in Introduction to American Government (fall 2021 to fall 2022) and 915 in Principles of Microeconomics (2022-23) at Georgia State University.","uncertainty":"Numeric final grade effect not significant; men in Microeconomics showed no effect; several coauthors work at the deploying university.","comparison_basis":"measured","source_urls":["https://www.nber.org/system/files/working_papers/w35397/w35397.pdf","https://www.nber.org/papers/w35397","https://doi.org/10.1016/j.econedurev.2026.102816"],"source_refs":["AIBD-0617-source-0003"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The chatbot service as a whole; the automated answering is not separated from staff-written messages.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Students randomly assigned to no chatbot.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Four percentage points from a control rate of 61 percent.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two or three text messages a week for a semester.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Final course grades and grade categories from university records.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Probability of earning an A or B.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Registry of Efficacy and Effectiveness Studies 8160 and 13760.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"1,568 students in Introduction to American Government (fall 2021 to fall 2022) and 915 in Principles of Microeconomics (2022-23) at Georgia State University.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Numeric final grade effect not significant; men in Microeconomics showed no effect; several coauthors work at the deploying university.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0617-reported-measure","evaluation_data":null},{"case_id":"AIBD-0048","case_url":"records/AIBD-0048.html","measure":"clinicians diagnosed an infant within five days","reported_result":"clinicians diagnosed an infant within five days","baseline":"one to two weeks for much more limited results from chromosomal microarray and four to six weeks for targeted gene panels","result":"less than five days","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://europepmc.org/article/PMC/PMC11234061","https://pmc.ncbi.nlm.nih.gov/articles/PMC9325884/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13705538/"],"source_refs":["AIBD-0048-source-0001","AIBD-0048-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: chromosomal microarray and targeted gene panels","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: one to two weeks for much more limited results from chromosomal microarray and four to six weeks for targeted gene panels. Result: less than five days.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"scored a homozygous ANTXR2 variant at 4.7, diagnosing an infant within five days","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0048-reported-measure","evaluation_data":null},{"case_id":"AIBD-0346","case_url":"records/AIBD-0346.html","measure":"Sensitivity, specificity and agreement (Cohen's kappa) for detecting damaged facilities.","reported_result":"would typically take weeks and now takes hours or minutes","baseline":"weeks","result":"hours or minutes","baseline_value":null,"result_value":null,"unit":null,"effect":"Microsoft model: sensitivity 42.86%, specificity 93.55%; sensitivity 100% for hospitals and 0% for dialysis centers; agreement with the Google model kappa 0.32. Aggregating damage over 0.125 km² cells raised sensitivity to 71.4%.","design":"other.","sample":"25 hospitals, 13 dialysis facilities and 454 pharmacies in Antakya, Kahramanmaraş and Iskenderun after the February 2023 earthquakes; sensitivity and specificity computed for the 38 hospitals and dialysis centers.","uncertainty":"Small sample with a wide interval (95% CI 9.90% to 81.59%); hospital reference based on imagery review; no reference data for pharmacies; the study does not examine whether responders used the maps.","comparison_basis":"assumed","source_urls":["https://www.technologyreview.com/2023/02/20/1068824/ai-actually-helpful-disaster-response-turkey-syria-earthquake/","https://www.microsoft.com/en-us/research/publication/turkey-earthquake-report/","https://arxiv.org/abs/1911.09296","https://www.directrelief.org/2025/12/after-disaster-ai-models-support-rapid-understanding-of-health-facility-damage/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12558478/"],"source_refs":["AIBD-0346-source-0009"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not measured. The study compares two models with reference data and does not compare them with a human-only damage survey.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Reference damage data, and a second model from Google Research and the WFP Innovation Accelerator.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Microsoft model: sensitivity 42.86%, specificity 93.55%; sensitivity 100% for hospitals and 0% for dialysis centers; agreement with the Google model kappa 0.32. Aggregating damage over 0.125 km² cells raised sensitivity to 71.4%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Model output for three cities in the February 2023 response.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective accuracy study. The authors overlaid each model's building damage output on geocoded health facilities and compared the result with reference damage data. For dialysis centers the reference was data from the Turkish Nephrological Institute; for hospitals it was the consensus of two reviewers who examined satellite images.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sensitivity, specificity and agreement (Cohen's kappa) for detecting damaged facilities.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Open-access article with methods and results tables.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"25 hospitals, 13 dialysis facilities and 454 pharmacies in Antakya, Kahramanmaraş and Iskenderun after the February 2023 earthquakes; sensitivity and specificity computed for the 38 hospitals and dialysis centers.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small sample with a wide interval (95% CI 9.90% to 81.59%); hospital reference based on imagery review; no reference data for pharmacies; the study does not examine whether responders used the maps.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0346-reported-measure","evaluation_data":null},{"case_id":"AIBD-0644","case_url":"records/AIBD-0644.html","measure":"Breast cancer detection rate and recall rate per 1,000.","reported_result":"found 6.7 cancers per 1,000 against 5.7 without it","baseline":"5.7 cancers per 1,000 women","result":"6.7 cancers per 1,000 women","baseline_value":null,"result_value":null,"unit":null,"effect":"6.7 vs 5.7 per 1,000 (+17.6%); recall 37.4 vs 38.3 per 1,000.","design":"controlled observational.","sample":"463,094 women aged 50 to 69 at 12 sites, July 2021 to February 2023; 119 radiologists.","uncertainty":"No randomization; reading-behavior bias reported; Vara funded and co-wrote the study.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11922743/","https://pubmed.ncbi.nlm.nih.gov/42617468/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13605948/"],"source_refs":["AIBD-0644-source-0003"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"1.0 more cancer per 1,000 women.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Double reading without AI.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"6.7 vs 5.7 per 1,000 (+17.6%); recall 37.4 vs 38.3 per 1,000.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"260,739 women read with AI support.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Screen-detected cancers and recalls from program records.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Breast cancer detection rate and recall rate per 1,000.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Anonymized dataset on Dryad; protocol online.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"463,094 women aged 50 to 69 at 12 sites, July 2021 to February 2023; 119 radiologists.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"No randomization; reading-behavior bias reported; Vara funded and co-wrote the study.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0644-reported-measure","evaluation_data":null},{"case_id":"AIBD-0353","case_url":"records/AIBD-0353.html","measure":"contributors using RapiD mapped tasks up to 12.4% slower","reported_result":"contributors using RapiD mapped tasks up to 12.4% slower","baseline":null,"result":"Contributors using Rapid mapped tasks up to 12.4% slower","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://wiki.openstreetmap.org/wiki/2023_Turkey_Earthquakes","https://web.archive.org/web/20230207114951/https://www.hotosm.org/disaster-services/turkiye-earthquakes-february-2023-activation/","https://web.archive.org/web/20230529125025/https://www.hotosm.org/disaster-services/turkiye-earthquakes-february-2023-size-up/","https://tasks.hotosm.org/projects/14262","https://wiki.openstreetmap.org/wiki/Rapid"],"source_refs":["AIBD-0353-source-0001","AIBD-0353-source-0002","AIBD-0353-source-0003","AIBD-0353-source-0004","AIBD-0353-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"contributors using RapiD mapped tasks up to 12.4% slower","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"contributors using RapiD mapped tasks up to 12.4% slower","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0353-reported-measure","evaluation_data":null},{"case_id":"AIBD-0243","case_url":"records/AIBD-0243.html","measure":"Share of trip time and distance on sidewalks.","reported_result":"The feedback group spent 22% less time and 20% less distance on sidewalks","baseline":"0.22 (mean fraction of trip time on sidewalks, no-feedback group, Table 1)","result":"0.17 (mean fraction of trip time on sidewalks, feedback group, Table 1)","baseline_value":null,"result_value":null,"unit":null,"effect":"22% less time and 20% less distance on sidewalks in the feedback group.","design":"randomized.","sample":"100 Spin e-scooters in Santa Monica, California; 456 analyzed trips (289 with feedback, 167 without) from 23 November 2022 to 14 February 2023.","uncertainty":"Small sample after theft led Spin to withdraw the camera scooters; incomplete trajectory data; detection accuracy not reported; the same camera classified the surface for both arms.","comparison_basis":"measured","source_urls":["https://findingspress.org/article/125892-ai-driven-feedback-systems-reduce-sidewalk-riding-on-shared-e-scooters"],"source_refs":["AIBD-0243-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"50 scooters with feedback switched off at random, riders unaware of assignment.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"22% less time and 20% less distance on sidewalks in the feedback group.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"On-board camera classification of the riding surface, with an event recorded at each change of surface.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Share of trip time and distance on sidewalks.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"100 Spin e-scooters in Santa Monica, California; 456 analyzed trips (289 with feedback, 167 without) from 23 November 2022 to 14 February 2023.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small sample after theft led Spin to withdraw the camera scooters; incomplete trajectory data; detection accuracy not reported; the same camera classified the surface for both arms.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0243-reported-measure","evaluation_data":null},{"case_id":"AIBD-0218","case_url":"records/AIBD-0218.html","measure":"regularly published on the court website within days of a Constitution Bench hearing","reported_result":"regularly published on the court website within days of a Constitution Bench hearing","baseline":"no permanent published record of oral arguments","result":"lawyer-checked transcripts for 9 of 25 Constitution Bench hearings by 18 May 2023, and regular publication within days of hearings by February 2024","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.livelaw.in/top-stories/supreme-court-ai-live-transcriptions-of-its-hearings-222108","https://www.scobserver.in/journal/scs-live-transcription-project-a-failed-experiment/","https://www.ndtv.com/india-news/supreme-court-using-ai-to-transcribe-arguments-by-constitution-benches-4647186","https://www.pib.gov.in/PressReleasePage.aspx?PRID=2226283%C2%AE=48&lang=2","https://www.scobserver.in/journal/rti-privacy-proportionality-empirical-evidenceand-ai-transcripts-key-themes-from-the-electoral-bonds-judgement/"],"source_refs":["AIBD-0218-source-0001","AIBD-0218-source-0002","AIBD-0218-source-0003","AIBD-0218-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"9 of 25 hearings had published transcripts by May 2023","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"9 of 25 hearings had published transcripts by May 2023","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0218-reported-measure","evaluation_data":null},{"case_id":"AIBD-0062","case_url":"records/AIBD-0062.html","measure":"cancer detection rate per 1,000 women, recall rate, reader workload.","reported_result":"Mia helped find 12% more cancers than routine practice","baseline":"106 cancers found by routine double reading among 10,889 women","result":"11 more cancers found when cases flagged by Mia but not recalled by the readers received an additional human review","baseline_value":null,"result_value":null,"unit":null,"effect":"11 additional cancers detected beyond the 106 found by routine screening; the primary workflow raised cancer detection by 10.4 percent (1 per 1,000) and kept the recall rate level.","design":"controlled observational.","sample":"NHS Grampian breast screening, Aberdeen, Scotland, in a program that invites women aged 50 to 71; 10,889 women screened between 27 February and 5 October 2023.","uncertainty":"One site and one AI tool. Three-year follow-up for interval cancers was not yet available, so the authors could not assess the true sensitivity of readers and AI. Triage workflows were simulated. Three authors were employed by Kheiron.","comparison_basis":"measured","source_urls":["https://www.nature.com/articles/s43018-026-01126-1","https://www.bbc.com/news/technology-68607059","https://www.abdn.ac.uk/news/22964/"],"source_refs":["AIBD-0062-source-0004"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The authors released the AI opinion to the arbitration readers only after the routine double reading decision. They argue that this design gives a true reflection of what the AI added to cancer detection.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"routine double reading with arbitration on the same women.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"11 additional cancers detected beyond the 106 found by routine screening; the primary workflow raised cancer detection by 10.4 percent (1 per 1,000) and kept the recall rate level.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mia v3 at a higher-sensitivity operating point, which flagged 12.4 percent of cases (1,345) for additional review.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mia read every eligible mammogram in the background. Cases that Mia flagged but routine double reading did not recall went to an additional human arbitration review. Cancers were confirmed by histology after surgery or biopsy. Triage workflows were simulated on the same data.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"cancer detection rate per 1,000 women, recall rate, reader workload.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The evaluation and statistical analysis plans are in the Supplementary Information, with bootstrapped confidence intervals as Supplementary Data. The R code is on GitHub and Zenodo. Images and data stay in the Grampian Data Safe Haven and are available on approved request.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"NHS Grampian breast screening, Aberdeen, Scotland, in a program that invites women aged 50 to 71; 10,889 women screened between 27 February and 5 October 2023.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One site and one AI tool. Three-year follow-up for interval cancers was not yet available, so the authors could not assess the true sensitivity of readers and AI. Triage workflows were simulated. Three authors were employed by Kheiron.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0062-reported-measure","evaluation_data":null},{"case_id":"AIBD-0285","case_url":"records/AIBD-0285.html","measure":"atrial fibrillation with rapid ventricular response at 123 bpm","reported_result":"atrial fibrillation with rapid ventricular response at 123 bpm","baseline":"24 hours of chest pressure and palpitations, with no diagnosis","result":"atrial fibrillation confirmed on ECG and sinus rhythm restored by cardioversion the next day","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10082624/"],"source_refs":["AIBD-0285-source-0001","AIBD-0285-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"atrial fibrillation with rapid ventricular response at 123 bpm","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"atrial fibrillation with rapid ventricular response at 123 bpm","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0285-reported-measure","evaluation_data":null},{"case_id":"AIBD-0138","case_url":"records/AIBD-0138.html","measure":null,"reported_result":null,"baseline":"the vet had no clue what it could be","result":"After more blood tests, \"the diagnosis was confirmed. GPT4 was right.\"","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.businessinsider.com/twitter-user-claims-gpt4-saved-dogs-life-vet-couldnt-diagnose-2023-3","https://twitter.com/peakcooper/status/1639716822680236032","https://nypost.com/2023/03/27/chatgpt-saved-my-dogs-life-after-vet-couldnt-diagnosis-it/","https://www.indiatoday.in/technology/news/story/chatgpt-saves-life-of-dog-diagnoses-problem-that-even-vet-could-not-identify-2352306-2023-03-28","https://web.archive.org/web/2023/https://openai.com/research/gpt-4"],"source_refs":["AIBD-0138-source-0002","AIBD-0138-source-0003","AIBD-0138-source-0004","AIBD-0138-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the first veterinarian, who found no cause after further tests and advised waiting","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: the vet had no clue what it could be. Result: After more blood tests, \"the diagnosis was confirmed. GPT4 was right.\".","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0138-reported-measure","evaluation_data":null},{"case_id":"AIBD-0116","case_url":"records/AIBD-0116.html","measure":"£60 parking fine; York City Council withdrew it","reported_result":"£60 parking fine; York City Council withdrew it","baseline":"A £60 fine notice issued by York City Council","result":"The notice withdrawn by the council","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://web.archive.org/web/2026/https://www.itv.com/news/calendar/2023-03-31/appeal-written-by-ai-chatbot-helps-student-overturn-parking-fine","https://www.itv.com/news/calendar/2023-03-31/appeal-written-by-ai-chatbot-helps-student-overturn-parking-fine","https://www.bbc.co.uk/news/uk-england-york-north-yorkshire-65126772","https://www.yorkshirepost.co.uk/read-this/student-overturns-ps60-parking-ticket-using-chatgpt-4088236","https://www.foxnews.com/world/how-ai-chatbot-allegedly-helped-student-terminate-parking-fine-very-relieved"],"source_refs":["AIBD-0116-source-0002","AIBD-0116-source-0001","AIBD-0116-source-0003","AIBD-0116-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"£60 parking fine; York City Council withdrew it","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"£60 parking fine; York City Council withdrew it","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0116-reported-measure","evaluation_data":null},{"case_id":"AIBD-0080","case_url":"records/AIBD-0080.html","measure":"confirmed early-stage lung cancer in five adults, and four had surgery","reported_result":"confirmed early-stage lung cancer in five adults, and four had surgery","baseline":"would likely have been missed or delayed","result":"early-stage lung cancer in all five cases","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC13058629/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12914670/"],"source_refs":["AIBD-0080-source-0001","AIBD-0080-source-0002","AIBD-0080-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: routine chest X-ray reading without the AI","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: would likely have been missed or delayed. Result: early-stage lung cancer in all five cases.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"CT/biopsy confirmed early-stage lung cancer in five adults; four had surgery","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0080-reported-measure","evaluation_data":null},{"case_id":"AIBD-0338","case_url":"records/AIBD-0338.html","measure":null,"reported_result":null,"baseline":"more than 80 percent of large partnership audits closed with no change to the return","result":"not yet measured; TIGTA reported in May 2025 that no examination selected by the model had closed","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.gao.gov/assets/gao-23-106020.pdf","https://www.irs.gov/newsroom/irs-announces-sweeping-effort-to-restore-fairness-to-tax-system-with-inflation-reduction-act-funding-new-compliance-efforts","https://www.gao.gov/blog/artificial-intelligence-may-help-irs-close-tax-gap","https://www.oversight.gov/sites/default/files/documents/reports/2025-05/2025308022fr.pdf"],"source_refs":["AIBD-0338-source-0001","AIBD-0338-source-0002","AIBD-0338-source-0003","AIBD-0338-source-0004","AIBD-0338-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0338-reported-measure","evaluation_data":null},{"case_id":"AIBD-0297","case_url":"records/AIBD-0297.html","measure":"Diagnostic yield and clinical utility per registration session; agreement between the algorithm and expert consensus.","reported_result":"Nelli flagged 707 possible seizure events, 247 unheard or unrecognized by carers","baseline":"247 were not heard or recognised by carers","result":"Nelli recorded 707 possible seizure events across the study cohort","baseline_value":null,"result_value":null,"unit":null,"effect":"Diagnostic yield in 11 of 20 sessions and clinical utility in 8 of 20; positive predictive value of 81.8% for sessions classified as epileptic; no significant difference between the algorithm and the experts.","design":"other.","sample":"20 patients with focal pharmacoresistant epilepsy referred for inpatient video-EEG monitoring at Linköping University Hospital, monitored at home for a median of 15.5 nights.","uncertainty":"Small exploratory sample of 20 patients; yield and utility were much lower for subtle focal motor seizures.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10160340/","https://pmc.ncbi.nlm.nih.gov/articles/PMC9480955/","https://europepmc.org/article/MED/41066202","https://pubmed.ncbi.nlm.nih.gov/39536364/","https://www.accessdata.fda.gov/cdrh_docs/pdf25/K251506.pdf"],"source_refs":["AIBD-0297-source-0006"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Clinical experts' consensus assessment of the captured recordings.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Diagnostic yield in 11 of 20 sessions and clinical utility in 8 of 20; positive predictive value of 81.8% for sessions classified as epileptic; no significant difference between the algorithm and the experts.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One home registration session per patient.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Clinical experts reviewed the captured video recordings and classified each registration session as diagnostic or not; clinical utility was judged against pre-specified measures.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Diagnostic yield and clinical utility per registration session; agreement between the algorithm and expert consensus.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"20 patients with focal pharmacoresistant epilepsy referred for inpatient video-EEG monitoring at Linköping University Hospital, monitored at home for a median of 15.5 nights.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small exploratory sample of 20 patients; yield and utility were much lower for subtle focal motor seizures.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0297-reported-measure","evaluation_data":null},{"case_id":"AIBD-0277","case_url":"records/AIBD-0277.html","measure":"BLEU4 scores of 36.52 and 37.47","reported_result":"BLEU4 scores of 36.52 and 37.47","baseline":"the translation memory baseline achieved a BLEU4 score of 23.51","result":"36.52 and 37.47 Best Bilingual Evaluation Understudy 4 (BLEU4) scores for C2E and T2E, respectively","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10153418/","https://pmc.ncbi.nlm.nih.gov/articles/PMC7592802/","https://arxiv.org/abs/2608.02609","https://web.archive.org/web/2026/https://www.timesofisrael.com/groundbreaking-ai-project-translates-5000-year-old-cuneiform-at-push-of-a-button/"],"source_refs":["AIBD-0277-source-0001","AIBD-0277-source-0002","AIBD-0277-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a translation memory baseline on the same test data","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: the translation memory baseline achieved a BLEU4 score of 23.51. Result: 36.52 and 37.47 Best Bilingual Evaluation Understudy 4 (BLEU4) scores for C2E and T2E, respectively.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"BLEU4 scores of 36.52 and 37.47","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0277-reported-measure","evaluation_data":null},{"case_id":"AIBD-0618","case_url":"records/AIBD-0618.html","measure":"Percentage-point learning gain from pre-test to post-test.","reported_result":"The ChatGPT group showed a 17.00 percent learning gain from pre-test to post-test.","baseline":"1.85 percent gain in the no-hint control (not significant)","result":"17.00 percent gain with ChatGPT hints","baseline_value":null,"result_value":null,"unit":null,"effect":"17.00 percent gain with ChatGPT hints against 1.85 percent with no hints and 11.62 percent with human-tutor hints.","design":"randomized.","sample":"274 Amazon Mechanical Turk workers with at least a high school degree, 28 February to 10 May 2023.","uncertainty":"Adult crowd workers; short sessions; three-item tests; 32 percent of ChatGPT hints failed quality checks before mitigation.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11125466/"],"source_refs":["AIBD-0618-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"ChatGPT hints produced a gain significant against no hints and similar to human-tutor hints.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"No-hint control and human-tutor hints.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"17.00 percent gain with ChatGPT hints against 1.85 percent with no hints and 11.62 percent with human-tutor hints.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One session of about 10 to 20 minutes.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Three-item pre-test and post-test on the assigned skill.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Percentage-point learning gain from pre-test to post-test.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Participant data at doi:10.6084/m9.figshare.23935269.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"274 Amazon Mechanical Turk workers with at least a high school degree, 28 February to 10 May 2023.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Adult crowd workers; short sessions; three-item tests; 32 percent of ChatGPT hints failed quality checks before mitigation.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0618-reported-measure","evaluation_data":null},{"case_id":"AIBD-0233","case_url":"records/AIBD-0233.html","measure":"detained a gang of 10 aged 15–22 behind eight accredited violent robberies","reported_result":"detained a gang of 10 aged 15–22 behind eight accredited violent robberies","baseline":"analizar cientos de carpetas para dar con potenciales vínculos","result":"el sistema realiza un filtro y propone un número acotado de ellos","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.t13.cl/noticia/nacional/operacion-influencer-inteligencia-artificial-clave-detencion-banda-encerronas-12-5-2023","https://www.fiscaliadechile.cl/actualidad/noticias/regionales/operacion-influencer-uso-de-ia-permitio-desarticular-banda-de","https://oecd.ai/en/incidents/2023-05-12-99fe","https://www.fiscaliadechile.cl/actualidad/noticias/en-viena-fiscal-heredia-obtiene-reconocimiento-en-prestigioso-concurso"],"source_refs":["AIBD-0233-source-0001","AIBD-0233-source-0002","AIBD-0233-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: prosecutors manually analyzing hundreds of case folders to find links","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: analizar cientos de carpetas para dar con potenciales vínculos. Result: el sistema realiza un filtro y propone un número acotado de ellos.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"detained a gang of 10 aged 15–22 behind eight accredited violent robberies","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0233-reported-measure","evaluation_data":null},{"case_id":"AIBD-0171","case_url":"records/AIBD-0171.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.media.enea.it/comunicati-e-news/archivio-anni/anno-2023/energia-enea-intelligenza-artificiale-per-evitare-blackout-elettrici-da-ondate-di-calore.html","https://oecd.ai/en/incidents/2023-05-24-c85b"],"source_refs":["AIBD-0171-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0171-reported-measure","evaluation_data":null},{"case_id":"AIBD-0550","case_url":"records/AIBD-0550.html","measure":null,"reported_result":null,"baseline":null,"result":"increased prevention effectiveness by 65%, avoided 96 homelessness cases, and avoided 44 temporary accommodation placements","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.gov.uk/government/publications/a-national-plan-to-end-homelessness/a-national-plan-to-end-homelessness","https://www.youtube.com/watch?v=Nz5fzLtzO_Q","https://www.nesta.org.uk/project-updates/how-local-authorities-are-using-data-to-solve-problems-like-homelessness/"],"source_refs":["AIBD-0550-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is between supported flagged households and a group that lacked support because of limited capacity; allocation was not described as randomized.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0550-reported-measure","evaluation_data":null},{"case_id":"AIBD-0652","case_url":"records/AIBD-0652.html","measure":"Weighted kappa against the consensus panel.","reported_result":"hepatocellular ballooning (0.563 against 0.414)","baseline":"Mean weighted kappa of unassisted reads","result":"Weighted kappa higher by 0.15 for ballooning and 0.12 for lobular inflammation","baseline_value":null,"result_value":null,"unit":null,"effect":"Lobular inflammation 0.419 against 0.297; ballooning 0.563 against 0.414; steatosis and fibrosis similar.","design":"controlled observational.","sample":"1,481 cases from four completed MASH trials (three phase 2 and one phase 3).","uncertainty":"Clinical meaning of the gains questioned by CHMP; locked model and one scanner model; study funded by PathAI and drug companies.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11750710/","https://www.ema.europa.eu/en/documents/other/qualification-opinion-artificial-intelligence-based-measurement-non-alcoholic-steatohepatitis-histology-liver-biopsies-determine-disease-activity-nash-mash-clinical-trials_en.pdf","https://www.ema.europa.eu/en/news/ema-qualifies-first-artificial-intelligence-tool-diagnose-inflammatory-liver-disease-mash-biopsy-samples"],"source_refs":["AIBD-0652-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Weighted kappa higher by 0.12 to 0.15 for inflammation and ballooning.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Individual manual reads of the same cases.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Lobular inflammation 0.419 against 0.297; ballooning 0.563 against 0.414; steatosis and fibrosis similar.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One AI-assisted read per case.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Weighted kappa of AI-assisted reads and of individual manual reads against panels of two expert pathologists with a tiebreaker.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Weighted kappa against the consensus panel.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Analysis code on GitHub (Path-AI/NASH_DDT_Manuscript).","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"1,481 cases from four completed MASH trials (three phase 2 and one phase 3).","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Clinical meaning of the gains questioned by CHMP; locked model and one scanner model; study funded by PathAI and drug companies.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0652-reported-measure","evaluation_data":null},{"case_id":"AIBD-0165","case_url":"records/AIBD-0165.html","measure":"all but one shot disrupted before it took control, and only two did after","reported_result":"all but one shot disrupted before it took control, and only two did after","baseline":"all shots except one (193266: low-βN reference shown in Fig. 3b) were disrupted","result":"only two (193277 and 193282) were disrupted","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.nature.com/articles/s41586-024-07024-9","https://engineering.princeton.edu/news/2024/02/21/engineers-use-ai-wrangle-fusion-power-grid","https://oecd.ai/en/incidents/2024-02-24-7d77"],"source_refs":["AIBD-0165-source-0001","AIBD-0165-source-0002","AIBD-0165-source-0003","AIBD-0165-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"DIII-D discharges in the same experiment session without RL control","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: all shots except one (193266: low-βN reference shown in Fig. 3b) were disrupted. Result: only two (193277 and 193282) were disrupted.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"before RL control all but one shot disrupted, after it only two did","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0165-reported-measure","evaluation_data":null},{"case_id":"AIBD-0666","case_url":"records/AIBD-0666.html","measure":"Time taken and output quality.","reported_result":"cut time on writing tasks by 40% and raised graded quality by 18%","baseline":"Control group time and graded quality","result":"40% less time, 18% higher quality","baseline_value":null,"result_value":null,"unit":null,"effect":"Time down 40%, quality up 18%.","design":"randomized.","sample":"453 college-educated professionals in an online experiment.","uncertainty":"Short designed tasks completed online; effects on workplace output not measured.","comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/37440646/","https://economics.mit.edu/sites/default/files/inline-files/Noy_Zhang_1.pdf"],"source_refs":["AIBD-0666-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"40% faster with 18% higher quality.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Participants without ChatGPT.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time down 40%, quality up 18%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"ChatGPT access for one writing task.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time on task and blinded grades from evaluators in the same occupations.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time taken and output quality.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Preregistered at the AEA RCT Registry (AEARCTR-0010882).","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"453 college-educated professionals in an online experiment.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Short designed tasks completed online; effects on workplace output not measured.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0666-reported-measure","evaluation_data":null},{"case_id":"AIBD-0146","case_url":"records/AIBD-0146.html","measure":"about 600 feet across with an Earth MOID of 0.0015 AU","reported_result":"about 600 feet across with an Earth MOID of 0.0015 AU","baseline":"ATLAS had observed 2022 SF289 three times on four separate nights, but never the requisite four times on one night to be identified as a new NEO","result":"It successfully combined fragments of data from all four nights and made the discovery.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.washington.edu/news/2023/07/31/heliolinc3d/","https://minorplanetcenter.net/mpec/K23/K23O26.html","https://minorplanetcenter.net/db_search/show_object?object_id=2022+SF289","https://www.hawaii.edu/news/2023/07/31/identify-potentially-hazardous-asteroid/","https://rubinobservatory.org/news/new-algorithm-asteroid","https://www.washington.edu/news/2026/04/02/rubin-observatory-11000-new-asteroids/","https://web.archive.org/web/20240109090840/https://www.nytimes.com/2023/08/05/science/space-asteroids-rubin-heliolinc3d.html"],"source_refs":["AIBD-0146-source-0001","AIBD-0146-source-0002","AIBD-0146-source-0003","AIBD-0146-source-0004","AIBD-0146-source-0005","AIBD-0146-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"conventional near-Earth-object discovery, which needs four observations of an object on one night","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: ATLAS had observed 2022 SF289 three times on four separate nights, but never the requisite four times on one night to be identified as a new NEO. Result: It successfully combined fragments of data from all four nights and made the discovery.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"about 600 feet across with an Earth MOID of 0.0015 AU","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0146-reported-measure","evaluation_data":null},{"case_id":"AIBD-0355","case_url":"records/AIBD-0355.html","measure":"kharif payouts in the three districts fell from Rs 217.91 crore to Rs 55.2 crore","reported_result":"kharif payouts in the three districts fell from Rs 217.91 crore to Rs 55.2 crore","baseline":"R217.91 crore","result":"R55.2 crore","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.downtoearth.org.in/agriculture/beyond-the-lens","https://www.pib.gov.in/PressReleaseIframePage.aspx?PRID=1941597","https://www.pib.gov.in/PressReleasePage.aspx?PRID=2104175","https://www.pib.gov.in/PressNoteDetails.aspx?id=157351&NoteId=157351&ModuleId=3","https://cdn.cseindia.org/attachments/0.70726700_1779095588_yestech.pdf","https://www.downtoearth.org.in/agriculture/can-technology-make-indias-crop-insurance-payouts-more-accurate"],"source_refs":["AIBD-0355-source-0001","AIBD-0355-source-0002","AIBD-0355-source-0003","AIBD-0355-source-0004","AIBD-0355-source-0005","AIBD-0355-source-0006","AIBD-0355-source-0007","AIBD-0355-source-0008"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"claims paid under manual crop-cutting experiments (2018)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: R217.91 crore. Result: R55.2 crore.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"kharif payouts in the three districts fell from Rs 217.91 crore to Rs 55.2 crore","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0355-reported-measure","evaluation_data":null},{"case_id":"AIBD-0276","case_url":"records/AIBD-0276.html","measure":"97% facial similarity with the Sistine Madonna","reported_result":"97% facial similarity with the Sistine Madonna","baseline":"thought by some experts to have been a Victorian copy","result":"found by the Bradford and Nottingham teams to be highly likely by Raphael, with 97% similarity between the two Madonnas","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.bbc.co.uk/news/uk-england-leeds-66291225","https://www.theartnewspaper.com/2023/07/26/clearly-a-copy-from-the-19th-centuryold-masters-scholars-reject-ai-attributed-raphael","https://news.artnet.com/art-world/ai-art-authentication-will-not-replace-humans-2340290","https://www.bradfordmuseums.org/whats-on/the-de-brecy-tondo","https://art-recognition.com/case-studies/raphael-the-brecy-tondo/","https://www.bradford.ac.uk/news/archive/2023/ai-study-shows-raphael-painting-was-not-entirely-the-masters-work.php"],"source_refs":["AIBD-0276-source-0001","AIBD-0276-source-0002","AIBD-0276-source-0003","AIBD-0276-source-0004","AIBD-0276-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"97% facial similarity with the Sistine Madonna","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"97% facial similarity with the Sistine Madonna","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0276-reported-measure","evaluation_data":null},{"case_id":"AIBD-0044","case_url":"records/AIBD-0044.html","measure":null,"reported_result":null,"baseline":"Tmem81 had no known role in fertilization, and biochemical attempts to find an IZUMO1–SPACA6 complex had limited success","result":"Tmem81 identified as part of a sperm complex with Izumo1 and Spaca6 and shown to be essential for male fertility in zebrafish and mice","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11052572/","https://europepmc.org/abstract/MED/39423812"],"source_refs":["AIBD-0044-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0044-reported-measure","evaluation_data":null},{"case_id":"AIBD-0118","case_url":"records/AIBD-0118.html","measure":"87% success with reinforcement learning and 27% without it","reported_result":"87% success with reinforcement learning and 27% without it","baseline":"27% successful force control without the reinforcement-learning agent","result":"87% successful force control with the reinforcement-learning agent","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://doi.org/10.1038/s41591-026-04498-0","https://patch.com/new-york/massapequa/inspirational-quadraplegic-massapequa-man-amazing-clinical-trial","https://msnewsroommedia.s3.amazonaws.com/mcp/northwell/neuralbypass/03-2024/3-Soundbites_Script%20%28New_Aug._2024%29/Double%20Neural%20Bypass%20Update%20-%20Bites%20Transcript.pdf","https://msnewsroommedia.s3.amazonaws.com/mcp/northwell/neuralbypass/02-2023/2023_Press_Release/DoubleNeuralBypass_Release.pdf","https://northwell.multimedia-newsroom.com/index.php/2026/07/16/paralyzed-man-has-feeling-lasting-movement-restored-in-hand/","https://clinicaltrials.gov/study/NCT03680872"],"source_refs":["AIBD-0118-source-0001","AIBD-0118-source-0002","AIBD-0118-source-0003","AIBD-0118-source-0004","AIBD-0118-source-0005","AIBD-0118-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Grasp-force control with the reinforcement-learning agent switched off, in the same participant","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 27% successful force control without the reinforcement-learning agent. Result: 87% successful force control with the reinforcement-learning agent.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"87% vs 27% successful grasp-force control with and without reinforcement learning","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0118-reported-measure","evaluation_data":null},{"case_id":"AIBD-0002","case_url":"records/AIBD-0002.html","measure":"Word error rate and words per minute.","reported_result":"248.3 hours of conversation over 8.4 months","baseline":"about 6.8 correct words per minute with expert listeners and 6.3 correct words per minute with a gyroscopic head mouse","result":"a self-paced 31.6 words per minute with a 2.5% word error rate in the final five Copy Task evaluation sessions","baseline_value":null,"result_value":null,"unit":null,"effect":"97.5% accuracy over 248 hours of conversation in 8.4 months; 2.5% word error rate at 31.6 words per minute in the final five Copy Task evaluation sessions.","design":"case study.","sample":"One 45-year-old man with ALS and severe dysarthria in the BrainGate2 trial, with sessions at his home over the 8.4 months after the July 2023 implant.","uncertainty":"One participant, and the authors are the developers. The Copy Task figures come from prompted sentences in research sessions.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11328962/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13375540/","https://www.sciencenews.org/article/brain-implants-restore-voices-injuries","https://www.sciencedaily.com/releases/2024/08/240814170129.htm","https://web.archive.org/web/20250605185915/https://www.nejm.org/doi/full/10.1056/NEJMp2407613"],"source_refs":["AIBD-0002-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"His usual communication with expert listeners (about 6.8 correct words per minute) and by head mouse (6.3 correct words per minute).","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"97.5% accuracy over 248 hours of conversation in 8.4 months; 2.5% word error rate at 31.6 words per minute in the final five Copy Task evaluation sessions.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"248 cumulative hours of conversational use over 8.4 months.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Word error rate and phoneme error rate as Levenshtein distance, chosen before data collection, on prompted sentences and in self-paced conversation; communication rate in words per minute.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Word error rate and words per minute.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One 45-year-old man with ALS and severe dysarthria in the BrainGate2 trial, with sessions at his home over the 8.4 months after the July 2023 implant.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One participant, and the authors are the developers. The Copy Task figures come from prompted sentences in research sessions.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0002-reported-measure","evaluation_data":null},{"case_id":"AIBD-0217","case_url":"records/AIBD-0217.html","measure":"Growth in raw test score from baseline to endline.","reported_result":"an effect size of 0.36 (Cohen's d, p < 0.001)","baseline":"control group growth score M = 2.12 (SD = 6.30) between baseline and endline","result":"treatment group growth score M = 5.13 (SD = 7.03), an effect size of 0.36 (Cohen's d)","baseline_value":null,"result_value":null,"unit":null,"effect":"Mean growth of 5.13 points in treatment against 2.12 in control, a difference of 3.01 points (p < 0.001), Cohen's d of 0.36.","design":"randomized.","sample":"11 Rising Academies schools in Ghana, five assigned to treatment and six to control at the school level; 637 students in grades 3 to 8 at baseline and 477 at endline.","uncertainty":"Assignment by school with only 11 schools; 25% attrition between tests; ceiling effects in higher grades; one year of data; one of the four authors lists Rising Academies, the developer, as affiliation, and a December 2024 trial registration lists the first author at Rising Academies as well.","comparison_basis":"measured","source_urls":["https://arxiv.org/pdf/2402.09809v2","https://arxiv.org/abs/2402.09809","https://www.risingacademies.com/risingnews/deepening-government-partnerships-and-responding-to-demand-in-ghana-1","https://www.socialscienceregistry.org/trials/14476"],"source_refs":["AIBD-0217-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The comparison isolates an extra hour a week of Rori practice against a normal study hall; it does not separate the chat tutor from the extra practice time.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Students in control schools who continued regular math instruction with no Rori access during study hall.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean growth of 5.13 points in treatment against 2.12 in control, a difference of 3.01 points (p < 0.001), Cohen's d of 0.36.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two 30-minute sessions a week over about eight months.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"The same 35-question math test (numeracy and algebra, grades 3 to 5 of the Global Proficiency Framework) given at baseline in February 2023 and at endline in August 2023; growth scores compared with a t-test and Cohen's d.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Growth in raw test score from baseline to endline.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"11 Rising Academies schools in Ghana, five assigned to treatment and six to control at the school level; 637 students in grades 3 to 8 at baseline and 477 at endline.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Assignment by school with only 11 schools; 25% attrition between tests; ceiling effects in higher grades; one year of data; one of the four authors lists Rising Academies, the developer, as affiliation, and a December 2024 trial registration lists the first author at Rising Academies as well.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0217-reported-measure","evaluation_data":null},{"case_id":"AIBD-0241","case_url":"records/AIBD-0241.html","measure":"Study 1: Claims per million miles. Study 2: Police-reportable crash involvements per million vehicle miles.","reported_result":"0.78 property-damage claims per million miles against 3.26, a 76% reduction","baseline":"1.11 claims per million miles (bodily injury); 3.26 claims per million miles (property damage)","result":"0.00 claims per million miles (bodily injury); 0.78 claims per million miles (property damage)","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: Bodily injury 0.00 against 1.11 claims per million miles (100% lower); property damage 0.78 against 3.26 (76% lower), with non-overlapping 95% confidence intervals. Study 2: 68% lower overall; 85% fewer single-vehicle crashes and 81% fewer injury crashes; 91% lower rate of rear-ending another vehicle.","design":"Study 1: controlled observational. Study 2: controlled observational.","sample":"Study 1: 3,868,506 rider-only miles in Phoenix and San Francisco from 1 January 2018 to 1 August 2023; human baseline from 2016 to 2021 claims and 125 billion miles. Study 2: 89 police-reportable Waymo crash involvements (64 in driverless operation) over about 50 million driverless miles in San Francisco, Phoenix, Los Angeles and Austin, 2021 to 2024; human drivers over about 222 billion miles in the same places and years.","uncertainty":"Study 1: Observational; the baseline uses the insured's registered zip code and cannot exclude freeway miles; small number of rider-only miles; Waymo staff co-wrote the paper. Study 2: Police-reportability coded by hand from narratives; 4% higher rate in Austin on a small sample; the authors call the coding method unsustainable as deployments grow.","comparison_basis":"measured","source_urls":["https://www.ebi.ac.uk/europepmc/webservices/rest/PMC11305169/fullTextXML","https://doi.org/10.1016/j.heliyon.2024.e34379","https://storage.googleapis.com/waymo-uploads/files/documents/safety/Comparison%20of%20Waymo%20and%20Human-Driven%20Vehicles%20at%2025M%20miles.pdf","https://waymo.com/blog/2024/12/new-swiss-re-study-waymo","https://doi.org/10.1080/15389588.2025.2499887","https://www.iihs.org/research-areas/bibliography/ref/2374"],"source_refs":["AIBD-0241-source-0001","AIBD-0241-source-0006"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: The automated driving system drove every rider-only mile, so the comparison is the system against human drivers in the same zip codes. Study 2: The comparison is the driverless system against human drivers on the same roads; it does not isolate any part of the system.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Zip-code-calibrated human-driver baseline from Swiss Re's portfolio. Study 2: Police-reported crash involvements of human drivers in the same areas and years.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: controlled observational. Study 2: controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Bodily injury 0.00 against 1.11 claims per million miles (100% lower); property damage 0.78 against 3.26 (76% lower), with non-overlapping 95% confidence intervals. Study 2: 68% lower overall; 85% fewer single-vehicle crashes and 81% fewer injury crashes; 91% lower rate of rear-ending another vehicle.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Rider-only operation with no human behind the wheel. Study 2: Driverless operation from 2021 to 2024.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: controlled observational. Study 2: controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Liability claims (property damage and bodily injury) per million miles for Waymo, adjudicated by its claims administrator, compared with a Swiss Re baseline of over 600,000 claims calibrated to the zip codes of Waymo's service areas. Study 2: Manual coding of NHTSA Standing General Order crash narratives to decide which crashes a person would report to police, then rates per million miles using Waymo's mileage and Federal Highway Administration mileage for humans.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Claims per million miles. Study 2: Police-reportable crash involvements per million vehicle miles.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 3,868,506 rider-only miles in Phoenix and San Francisco from 1 January 2018 to 1 August 2023; human baseline from 2016 to 2021 claims and 125 billion miles. Study 2: 89 police-reportable Waymo crash involvements (64 in driverless operation) over about 50 million driverless miles in San Francisco, Phoenix, Los Angeles and Austin, 2021 to 2024; human drivers over about 222 billion miles in the same places and years.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Observational; the baseline uses the insured's registered zip code and cannot exclude freeway miles; small number of rider-only miles; Waymo staff co-wrote the paper. Study 2: Police-reportability coded by hand from narratives; 4% higher rate in Austin on a small sample; the authors call the coding method unsustainable as deployments grow.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0241-reported-measure","evaluation_data":null},{"case_id":"AIBD-0667","case_url":"records/AIBD-0667.html","measure":"Minutes per task and grade.","reported_result":"cut time on four legal tasks by 11.8% to 32.1%","baseline":"69.72 minutes on average for the contract drafting task","result":"47.59 minutes on average with GPT-4 (32.1% less)","baseline_value":null,"result_value":null,"unit":null,"effect":"Time down 24.1% (complaint), 32.1% (contract), 21.1% (handbook), 11.8% (memo); grade gain significant only for the contract.","design":"randomized.","sample":"60 University of Minnesota law students, 240 task completions.","uncertainty":"Students in place of practicing lawyers; small sample; designed tasks.","comparison_basis":"measured","source_urls":["https://web.archive.org/web/2025id_/https://scholarship.law.umn.edu/cgi/viewcontent.cgi?article=1002&context=minnlrev","https://doi.org/10.24926/265535.4225"],"source_refs":["AIBD-0667-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Faster on all four tasks.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Same tasks without GPT-4.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time down 24.1% (complaint), 32.1% (contract), 21.1% (handbook), 11.8% (memo); grade gain significant only for the contract.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"GPT-4 access after several hours of training.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Self-tracked time on task and blind grading on a 4.0 scale.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Minutes per task and grade.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Task materials reproduced in the article appendices.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"60 University of Minnesota law students, 240 task completions.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Students in place of practicing lawyers; small sample; designed tasks.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0667-reported-measure","evaluation_data":null},{"case_id":"AIBD-0631","case_url":"records/AIBD-0631.html","measure":"Sensitivity, specificity and positive predictive value.","reported_result":"a sensitivity of 68.4% and a specificity of 99.1%","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":"Sensitivity 68.4%, specificity 99.1%, PPV 13.7% at a prevalence of 0.2%.","design":"other.","sample":"3,196 ECGs (291 with and 2,905 without hypertrophic cardiomyopathy) from three Massachusetts hospitals, 2017 to 2022.","uncertainty":"Retrospective; one cardiologist set most labels; about one in three patients with the disease was missed.","comparison_basis":null,"source_urls":["https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN230003.pdf"],"source_refs":["AIBD-0631-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Cardiologist review of the chart and imaging.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sensitivity 68.4%, specificity 99.1%, PPV 13.7% at a prevalence of 0.2%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One ECG per case.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective standalone accuracy study against cardiologist chart and imaging review.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sensitivity, specificity and positive predictive value.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"None reported.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"3,196 ECGs (291 with and 2,905 without hypertrophic cardiomyopathy) from three Massachusetts hospitals, 2017 to 2022.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective; one cardiologist set most labels; about one in three patients with the disease was missed.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0631-reported-measure","evaluation_data":null},{"case_id":"AIBD-0224","case_url":"records/AIBD-0224.html","measure":"311 probable identifications of previously unknown victims, 14 positive contacts","reported_result":"311 probable identifications of previously unknown victims, 14 positive contacts","baseline":"formerly cold cases","result":"there are now fresh leads","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.nbcnews.com/news/investigations/homeland-security-identifies-311-child-victims-sexual-exploitation-rcna97357","https://www.ice.gov/news/releases/hsi-partners-launch-first-us-based-international-victim-identification-surge","https://www.abc.net.au/news/2023-08-24/qld-facial-recognition-child-abuse-artificial-intelligence/102761618","https://www.forbes.com/sites/maryroeloffs/2023/08/09/dhs-says-it-identified-311-new-child-sexual-exploitation-victims-after-forbes-reports-ai-facial-recognition-push/","https://www.crikey.com.au/2024/01/25/clearview-ai-australian-police-operation-renewed-hope/"],"source_refs":["AIBD-0224-source-0001","AIBD-0224-source-0002","AIBD-0224-source-0003","AIBD-0224-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the same abuse-material series as previously reviewed, which had left the cases cold","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: formerly cold cases. Result: there are now fresh leads.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"311 probable identifications of unknown child victims, 14 positive contacts","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0224-reported-measure","evaluation_data":null},{"case_id":"AIBD-0120","case_url":"records/AIBD-0120.html","measure":"Cal Fire held the verified fire under a quarter acre","reported_result":"Cal Fire held the verified fire under a quarter acre","baseline":"not stated","result":"the AI beats 911 calls about 40 percent of the time","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://english.aawsat.com/node/4540976","https://phys.org/news/2023-09-california-firefighters-ai-wildfires.html","https://www.france24.com/en/live-news/20230912-california-firefighters-use-ai-to-battle-wildfires","https://www.omanobserver.om/article/1142729/opinion/firefighters-use-ai-to-battle-wildfires","https://alertcalifornia.org/cal-fire-and-university-of-california-san-diegos-alertcalifornia-program-join-forces-to-enhance-wildfire-response-with-artificial-intelligence-implementation-trial/","https://time.com/collection/best-inventions-2023/6327137/alertcalifornia-ai-wildfire-detector/","https://today.ucsd.edu/story/alertcalifornia-and-cal-fires-fire-detection-ai-program-named-one-of-times-best-inventions-of-2023","https://timesofsandiego.com/tech/2023/10/24/time-magazine-californias-ai-wildfire-detection-system-a-top-invention-of-2023/","https://alertcalifornia.org/technology/","https://www.latimes.com/california/story/2023-08-24/can-ai-fight-wildfires-california-is-counting-on-it","https://www.firerescue1.com/artificial-intelligence/over-1-000-ai-cameras-help-spot-calif-wildfires-before-the-first-911-call","https://www.ap.org/news-highlights/spotlights/2026/states-across-the-wildfire-prone-western-us-are-using-ai-for-early-detection/"],"source_refs":["AIBD-0120-source-0001","AIBD-0120-source-0002","AIBD-0120-source-0003","AIBD-0120-source-0004","AIBD-0120-source-0005","AIBD-0120-source-0006","AIBD-0120-source-0007","AIBD-0120-source-0008","AIBD-0120-source-0009"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"911 calls from the public as the first notice of a fire","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: not stated. Result: the AI beats 911 calls about 40 percent of the time.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Cal Fire held the verified fire under a quarter acre","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0120-reported-measure","evaluation_data":null},{"case_id":"AIBD-0085","case_url":"records/AIBD-0085.html","measure":"identified at least 247 similar appeals in the controversy behind Tema 1.261","reported_result":"identified at least 247 similar appeals in the controversy behind Tema 1.261","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://noticias.stf.jus.br/postsnoticias/inteligencia-artificial-ajuda-stf-a-identificar-processos-que-discutem-questoes-semelhantes/","https://portal.stf.jus.br/jurisprudenciaRepercussao/tema.asp?num=1261","https://portal.stf.jus.br/noticias/verNoticiaDetalhe.asp?idConteudo=471331","https://portal.stf.jus.br/jurisprudenciaRepercussao/verPronunciamento.asp?pronunciamento=10530534","https://noticias.stf.jus.br/postsnoticias/ministra-rosa-weber-lanca-robo-vitoria-para-agrupamento-e-classificacao-de-processos/"],"source_refs":["AIBD-0085-source-0001","AIBD-0085-source-0002","AIBD-0085-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"identified at least 247 similar appeals in the controversy behind Tema 1.261","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"identified at least 247 similar appeals in the controversy behind Tema 1.261","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0085-reported-measure","evaluation_data":null},{"case_id":"AIBD-0119","case_url":"records/AIBD-0119.html","measure":"About 40 ml of clot came out in just over 30 minutes","reported_result":"About 40 ml of clot came out in just over 30 minutes","baseline":"nếu mổ theo phương pháp truyền thống, nguy cơ di chứng cao (high risk of sequelae with the traditional method)","result":"khi được mổ lấy máu tụ tỉnh thức bằng robot AI, ông có thể song ca cùng bác sĩ ngay khi đang mổ (awake robot-assisted clot removal while talking and singing)","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://tamanhhospital.vn/tin-tuc/nguoi-benh-hat-trong-luc-duoc-robot-ai-mo-nao/","https://vietnamnet.vn/nguoi-benh-hat-trong-luc-duoc-robot-ai-mo-nao-2177047.html","https://baodautu.vn/tin-moi-ve-y-te-ngay-148-nguoi-benh-hat-trong-luc-duoc-robot-ai-mo-nao-d196299.html","https://tamanhhospital.vn/tin-tuc/ra-mat-robot-mo-nao-duy-nhat-viet-nam-tai-bvdk-tam-anh-tphcm/","https://www.synaptivemedical.com/news/synaptive-medical-announces-product-update-to-flagship-surgical-technology-modus-v-tm-automated-robotic-microscope","https://pmc.ncbi.nlm.nih.gov/articles/PMC12269361/"],"source_refs":["AIBD-0119-source-0001","AIBD-0119-source-0002","AIBD-0119-source-0003","AIBD-0119-source-0004","AIBD-0119-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: conventional (traditional) brain surgery under general anesthesia","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: nếu mổ theo phương pháp truyền thống, nguy cơ di chứng cao (high risk of sequelae with the traditional method). Result: khi được mổ lấy máu tụ tỉnh thức bằng robot AI, ông có thể song ca cùng bác sĩ ngay khi đang mổ (awake robot-assisted clot removal while talking and singing).","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"About 40 ml of clot came out in just over 30 minutes","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0119-reported-measure","evaluation_data":null},{"case_id":"AIBD-0313","case_url":"records/AIBD-0313.html","measure":"contain the Trotter Fire at 52 acres","reported_result":"contain the Trotter Fire at 52 acres","baseline":"nearly 4,000 acres","result":"52 acres","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://unlocked.microsoft.com/alertcalifornia-ai-wildfire-detection/","https://web.archive.org/web/20260505035356/https://www.fire.ca.gov/incidents/2023/8/14/trotter/","https://bakersfieldnow.com/news/local/kern-county-fire-department-orders-evacuation-in-bear-valley-springs-due-to-vegetation-fires-kern-county","https://alertcalifornia.org/new-collaboration-expands-capabilities-and-enhances-natural-disaster-resilience/"],"source_refs":["AIBD-0313-source-0001","AIBD-0313-source-0002","AIBD-0313-source-0003","AIBD-0313-source-0004","AIBD-0313-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the fire's estimated spread without AI-powered imaging, as estimated by ALERTCalifornia founder Neal Driscoll","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: nearly 4,000 acres. Result: 52 acres.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"contain the Trotter Fire at 52 acres","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0313-reported-measure","evaluation_data":null},{"case_id":"AIBD-0479","case_url":"records/AIBD-0479.html","measure":null,"reported_result":null,"baseline":"four working days per test","result":"eight minutes per test, with 650 percent more samples examined each day","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://techmgzn.com/dubai-municipality-to-test-construction-materials-using-ai-robots/","https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0479-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The submission describes a comparison of manual and automated cement testing using 60 dummy samples in Dubai under BS EN 196 procedures.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0479-reported-measure","evaluation_data":null},{"case_id":"AIBD-0341","case_url":"records/AIBD-0341.html","measure":"Mean track error, detection rate, intensity error and bias.","reported_result":"about nine days in advance that Hurricane Lee would make landfall in Nova Scotia","baseline":"only locked in on Nova Scotia about six days in advance","result":"about nine days in advance","baseline_value":null,"result_value":null,"unit":null,"effect":"GraphCast had mean track errors up to 13% smaller than the next best physics-based model (ECMWF) at 96 hours and later.","design":"controlled observational.","sample":"Northern Hemisphere tropical cyclones, May to November 2023; four open AI models run four times a day from GFS initial conditions, with a 2022 to 2023 comparison from IFS initial conditions.","uncertainty":"The sample is one season; the runs used GFS initial conditions and not the ECMWF-hosted live run, and the authors report that IFS initial conditions improved AI track forecasts significantly; the authors do not verify the Lee landfall forecast.","comparison_basis":"assumed","source_urls":["https://deepmind.google/discover/blog/graphcast-ai-model-for-faster-and-more-accurate-global-weather-forecasting/","https://web.archive.org/web/20231114164801/https://www.technologyreview.com/2023/11/14/1083366/google-deepminds-weather-ai-can-forecast-extreme-weather-quicker-and-more-accurately/","https://www.nhc.noaa.gov/data/tcr/AL132023_Lee.pdf","https://www.ecmwf.int/en/newsletter/177/editorial/aifs-launched","https://arxiv.org/abs/2212.12794","https://gmd.copernicus.org/articles/17/7915/2024/","https://journals.ametsoc.org/view/journals/aies/4/4/AIES-D-24-0085.1.xml"],"source_refs":["AIBD-0341-source-0009"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not statistically significant at long lead times; GraphCast errors were larger than ECMWF's at some early lead times.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Operational GFS and ECMWF models and the NHC official forecasts, on homogeneous samples.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"GraphCast had mean track errors up to 13% smaller than the next best physics-based model (ECMWF) at 96 hours and later.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"NHC verification procedures applied to 7-day track and intensity forecasts, compared with NHC and JTWC best tracks.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean track error, detection rate, intensity error and bias.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Northern Hemisphere tropical cyclones, May to November 2023; four open AI models run four times a day from GFS initial conditions, with a 2022 to 2023 comparison from IFS initial conditions.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The sample is one season; the runs used GFS initial conditions and not the ECMWF-hosted live run, and the authors report that IFS initial conditions improved AI track forecasts significantly; the authors do not verify the Lee landfall forecast.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0341-reported-measure","evaluation_data":null},{"case_id":"AIBD-0122","case_url":"records/AIBD-0122.html","measure":"found over 75 kilograms of narcotics","reported_result":"found over 75 kilograms of narcotics","baseline":"An officer would have had no cause to give the car more than a normal stop.","result":"The alert led the officer to refer the car to secondary inspection, where officers found over 75 kilograms of narcotics.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.congress.gov/118/meeting/house/116358/witnesses/HHRG-118-GO12-Wstate-HysenE-20230914.pdf","https://www.dhs.gov/ai","https://www.washingtonexaminer.com/policy/technology/mayorkas-dhs-ai-fentanyl-border-kidnapped-children","https://www.nextgov.com/artificial-intelligence/2023/11/dhs-aims-lead-defense-against-adversarial-ai/392321/","https://www.dhs.gov/sites/default/files/2024-03/24_0315_ocio_roadmap_artificialintelligence-ciov3-signed-508.pdf"],"source_refs":["AIBD-0122-source-0002","AIBD-0122-source-0003","AIBD-0122-source-0004","AIBD-0122-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"review found 75 kg of narcotics and the driver was arrested","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"review found 75 kg of narcotics and the driver was arrested","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0122-reported-measure","evaluation_data":null},{"case_id":"AIBD-0665","case_url":"records/AIBD-0665.html","measure":"Tasks completed, speed and quality score.","reported_result":"completed 12.2% more tasks, finished 25.1% more quickly","baseline":"Control group task completion, speed and quality","result":"12.2% more tasks, 25.1% faster, quality rated more than 40% higher","baseline_value":null,"result_value":null,"unit":null,"effect":"12.2% more tasks, 25.1% faster, more than 40% higher quality inside the frontier; 19 points lower correctness outside it.","design":"randomized.","sample":"758 BCG consultants, two experiments, 18 tasks inside the frontier and one outside.","uncertainty":"Designed tasks; one firm; one model version; BCG co-authors; lower correctness on the task outside the frontier.","comparison_basis":"measured","source_urls":["https://web.archive.org/web/20250102233043id_/https://www.hbs.edu/ris/Publication%20Files/24-013_d9b45b68-9e74-42d6-a1c6-c72fb70c7282.pdf","https://doi.org/10.1287/orsc.2025.21838"],"source_refs":["AIBD-0665-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Gains of 43% for below-average and 17% for above-average consultants.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Consultants without AI.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"12.2% more tasks, 25.1% faster, more than 40% higher quality inside the frontier; 19 points lower correctness outside it.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"GPT-4 access during the experimental task, with or without a prompt overview.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Subtasks completed, time to completion and graded quality of outputs.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Tasks completed, speed and quality score.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Preregistered on OSF; data not public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"758 BCG consultants, two experiments, 18 tasks inside the frontier and one outside.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Designed tasks; one firm; one model version; BCG co-authors; lower correctness on the task outside the frontier.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0665-reported-measure","evaluation_data":null},{"case_id":"AIBD-0034","case_url":"records/AIBD-0034.html","measure":null,"reported_result":null,"baseline":"Before GenAI, I would often send it to my mom or friend to check for me.","result":"Now I Ask ChatGPT","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://arxiv.org/html/2507.00286v1","https://pmc.ncbi.nlm.nih.gov/articles/PMC12338113/","https://pure.psu.edu/en/publications/before-i-asked-my-mom-now-i-ask-chatgpt-visual-privacy-management/","https://web.archive.org/web/2025/https://dl.acm.org/doi/10.1145/3663548.3675631","https://tsjournal.org/index.php/jots/article/view/340"],"source_refs":["AIBD-0034-source-0001","AIBD-0034-source-0002","AIBD-0034-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: asking a sighted family member or friend to check a photo","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: Before GenAI, I would often send it to my mom or friend to check for me. Result: Now I Ask ChatGPT.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0034-reported-measure","evaluation_data":null},{"case_id":"AIBD-0266","case_url":"records/AIBD-0266.html","measure":"median spelling rate of 10.2 characters per minute","reported_result":"median spelling rate of 10.2 characters per minute","baseline":"a spelling rate of 1.8 letters per minute with a latency of at least 1 s per click","result":"a spelling rate of 10.2 CCPM and 0.68 s latency","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11494178/","https://doi.org/10.21203/rs.3.rs-3158792/v1","https://doi.org/10.1088/1741-2552/ae7694","https://web.archive.org/web/2026/https://www.medrxiv.org/content/10.1101/2025.09.26.25335804v1"],"source_refs":["AIBD-0266-source-0001","AIBD-0266-source-0002","AIBD-0266-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the Utrecht fully implanted switch-scan BCI reported by Vansteensel et al. (four contacts over hand motor cortex)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: a spelling rate of 1.8 letters per minute with a latency of at least 1 s per click. Result: a spelling rate of 10.2 CCPM and 0.68 s latency.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"median spelling rate of 10.2 characters per minute","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0266-reported-measure","evaluation_data":null},{"case_id":"AIBD-0654","case_url":"records/AIBD-0654.html","measure":"Contouring time; clinical acceptability of contours.","reported_result":"Reported time savings ranged from 3 to 80 minutes","baseline":"Manual contouring time","result":"3 to 80 minutes saved in studies; experts estimated 10 to 30 minutes","baseline_value":null,"result_value":null,"unit":null,"effect":"AI contouring consistently quicker; contours generally similar.","design":"controlled observational.","sample":"79 studies; data extracted from 15 priority studies, 8 prospective.","uncertainty":"Some time savings excluded review time; small structures often needed major edits; limited cost evidence.","comparison_basis":"measured","source_urls":["https://www.nice.org.uk/guidance/htg695/chapter/3-Committee-discussion","https://www.nice.org.uk/guidance/htg695/chapter/1-Recommendations"],"source_refs":["AIBD-0654-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"3 to 80 minutes saved.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Manual contouring.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"AI contouring consistently quicker; contours generally similar.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Radiotherapy plans contoured with AI and reviewed by staff.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Review of contour quality ratings and contouring time with AI and with manual methods.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Contouring time; clinical acceptability of contours.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"External assessment group report on the NICE website.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"79 studies; data extracted from 15 priority studies, 8 prospective.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Some time savings excluded review time; small structures often needed major edits; limited cost evidence.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0654-reported-measure","evaluation_data":null},{"case_id":"AIBD-0287","case_url":"records/AIBD-0287.html","measure":null,"reported_result":null,"baseline":null,"result":"Within seconds, the application confirmed the diagnostic hypothesis","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://journals.viamedica.pl/international_maritime_health/article/view/99319"],"source_refs":["AIBD-0287-source-0002","AIBD-0287-source-0001","AIBD-0287-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: no electrocardiograph on board the ship","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Result: Within seconds, the application confirmed the diagnostic hypothesis.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0287-reported-measure","evaluation_data":null},{"case_id":"AIBD-0545","case_url":"records/AIBD-0545.html","measure":null,"reported_result":null,"baseline":"About six minutes per call wrap-up, implied by a two-minute saving equal to 33%","result":"About four minutes per call wrap-up","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["http://web.archive.org/web/20250915212319/https://www.local.gov.uk/case-studies/buckinghamshire-council-early-adoption-microsoft-copilot","https://www.local.gov.uk/case-studies/buckinghamshire-council-early-adoption-microsoft-copilot"],"source_refs":["AIBD-0545-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0545-reported-measure","evaluation_data":null},{"case_id":"AIBD-0568","case_url":"records/AIBD-0568.html","measure":"Planning officer time per consultation spent summarizing and reporting feedback.","reported_result":"Average saving of up to 85 per cent of officer time.","baseline":"Officer time needed to summarize and report on the February 2022 Cotswold consultation","result":"An average saving of up to 85 per cent of officer time on the West Oxfordshire consultation of August to October 2023","baseline_value":null,"result_value":null,"unit":null,"effect":"Average saving of up to 85 per cent of officer time.","design":"before/after.","sample":"One West Oxfordshire Local Plan consultation (August to October 2023) against one Cotswold consultation (February 2022).","uncertainty":"The two consultations were held by different councils at different times; the case study says that savings would vary with the type and volume of responses. Raw hours and the method of timing are not published.","comparison_basis":"measured","source_urls":["https://www.localdigital.gov.uk/case-studies/exploring-how-artificial-intelligence-ai-can-reduce-the-time-taken-to-report-on-consultation-feedback/","https://www.local.gov.uk/case-studies/west-oxfordshire-and-cotswold-district-councils-summarising-planning-engagement"],"source_refs":["AIBD-0568-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Cotswold consultation of February 2022, processed without the AI tools.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Average saving of up to 85 per cent of officer time.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Officer time to summarize and report on consultation feedback, compared between two consultations.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Planning officer time per consultation spent summarizing and reporting feedback.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One West Oxfordshire Local Plan consultation (August to October 2023) against one Cotswold consultation (February 2022).","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The two consultations were held by different councils at different times; the case study says that savings would vary with the type and volume of responses. Raw hours and the method of timing are not published.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0568-reported-measure","evaluation_data":null},{"case_id":"AIBD-0198","case_url":"records/AIBD-0198.html","measure":"mean fish activity (%) in the two months before and after the activity rise.","reported_result":"fish activity rose from 25.6% to 43.6% (Farm A) and 24.9% to 32.6% (Farm B)","baseline":"mean activity 25.6% at Farm A","result":"mean activity 43.6% at Farm A in the two months after","baseline_value":null,"result_value":null,"unit":null,"effect":"activity rose from 25.6% to 43.6% at Farm A and from 24.9% to 32.6% at Farm B; proliferative gill disease was the main driver in the mixed model.","design":"before/after.","sample":"Two commercial Atlantic salmon marine farms in the Western Isles of Scotland, 13 and 10 cages, May to October 2023.","uncertainty":"Three authors work for the vendor. The algorithm's speed and cohesion components were not validated separately, and other causes of the behavior change could not be ruled out.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12369403/","https://doi.org/10.1186/s44365-025-00020-8"],"source_refs":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12369403/"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"not measured; no comparison with detection by farm staff alone.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"the same cages in the two months before the rise.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"activity rose from 25.6% to 43.6% at Farm A and from 24.9% to 32.6% at Farm B; proliferative gill disease was the main driver in the mixed model.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"algorithm applied to all cages; four or five extra cameras in one cage per farm.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Algorithm activity output from feed cameras compared with weekly gill scores (amoebic and proliferative gill disease), specific feeding rate and mortality, with a mixed model for drivers of activity.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"mean fish activity (%) in the two months before and after the activity rise.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"raw data in the DataSTORRE repository, http://hdl.handle.net/11667/241.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two commercial Atlantic salmon marine farms in the Western Isles of Scotland, 13 and 10 cages, May to October 2023.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Three authors work for the vendor. The algorithm's speed and cohesion components were not validated separately, and other causes of the behavior change could not be ruled out.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0198-reported-measure","evaluation_data":null},{"case_id":"AIBD-0201","case_url":"records/AIBD-0201.html","measure":"helped the farm increase survival of lion-head geese by 30 percent","reported_result":"helped the farm increase survival of lion-head geese by 30 percent","baseline":null,"result":"survival rate of lion-head geese increased by 30 percent","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://english.news.cn/20231013/3c56fe0ec8b74246a934ca2ad09ac3ea/c.html","https://news.cgtn.com/news/2023-10-14/AI-program-developed-by-Chinese-college-students-benefits-goose-farm-1nTtBT0TBlK/index.html","https://oecd.ai/en/incidents/2023-10-13-d63c","https://www.nature.com/articles/s41597-024-03776-1"],"source_refs":["AIBD-0201-source-0001","AIBD-0201-source-0002","AIBD-0201-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the farm's survival rate before the AI program (no baseline figure is given)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Result: survival rate of lion-head geese increased by 30 percent.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"helped the farm increase survival of lion-head geese by 30 percent","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0201-reported-measure","evaluation_data":null},{"case_id":"AIBD-0090","case_url":"records/AIBD-0090.html","measure":"Track and intensity forecast error out to 7 days; change in NHC consensus track error.","reported_result":"adding all four improved the NHC track consensus by 11% at 120 h","baseline":"consensus track errors without the AIWP models; official five-day track forecasts improving by about 2% per year since 2001","result":"consensus track errors reduced by up to 11% at the longer time periods","baseline_value":null,"result_value":null,"unit":null,"effect":"AI track errors comparable to the best operational models; consensus track error 11% lower at 120 h with all four AI models added, significant at 72 to 120 h; intensity forecasts had no skill.","design":"controlled observational.","sample":"Northern Hemisphere tropical cyclones, May to November 2023; four open-source AI models run four times a day from GFS initial conditions.","uncertainty":"One season; the AI models almost always weakened storms, a substantial low intensity bias; track forecasts improved with ECMWF initial conditions in a preliminary test.","comparison_basis":"measured","source_urls":["https://arxiv.org/abs/2409.06735","https://journals.ametsoc.org/view/journals/aies/4/4/AIES-D-24-0085.1.xml","https://deepmind.google/blog/graphcast-ai-model-for-faster-and-more-accurate-global-weather-forecasting/","https://arxiv.org/abs/2212.12794","https://www.technologyreview.com/2023/11/14/1083366/google-deepminds-weather-ai-can-forecast-extreme-weather-quicker-and-more-accurately/","https://www.ecmwf.int/en/newsletter/177/editorial/aifs-launched","https://arxiv.org/abs/2312.02658","https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf"],"source_refs":["AIBD-0090-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"11% reduction in consensus track error at 120 h.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Operational physics-based models, official NHC forecasts and the consensus without the AI models.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"AI track errors comparable to the best operational models; consensus track error 11% lower at 120 h with all four AI models added, significant at 72 to 120 h; intensity forecasts had no skill.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"CIRA ran the models in near-real time and built a database of retrospective runs; NHC first brought AI models into real-time operations in 2025.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"NHC verification procedures, including conversion to early models and consensus experiments with and without the AI models.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Track and intensity forecast error out to 7 days; change in NHC consensus track error.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Northern Hemisphere tropical cyclones, May to November 2023; four open-source AI models run four times a day from GFS initial conditions.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One season; the AI models almost always weakened storms, a substantial low intensity bias; track forecasts improved with ECMWF initial conditions in a preliminary test.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0090-reported-measure","evaluation_data":null},{"case_id":"AIBD-0528","case_url":"records/AIBD-0528.html","measure":null,"reported_result":null,"baseline":"3 min","result":"15 s","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-1-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-891","https://www.thenationalnews.com/news/uae/2024/11/05/sheikh-mohammed-honours-winners-of-the-uae-ai-award/"],"source_refs":["AIBD-0528-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0528-reported-measure","evaluation_data":null},{"case_id":"AIBD-0417","case_url":"records/AIBD-0417.html","measure":null,"reported_result":null,"baseline":"Waste generated in public places was not only difficult to separate manually afterward but was also often collected already contaminated, rendering separation meaningless.","result":"Recyclables separated from general waste at the point of disposal","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-1-PDF-E.pdf","https://www.sedaily.com/NewsView/29X575U740"],"source_refs":["AIBD-0417-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0417-reported-measure","evaluation_data":null},{"case_id":"AIBD-0379","case_url":"records/AIBD-0379.html","measure":null,"reported_result":null,"baseline":"A tracked robot could not climb the bus steps","result":"The quadruped climbed into the bus and stood between the man and the suspected gun","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://cityclerk.lacity.org/onlinedocs/2022/22-1438-s1_rpt_bpc_3-21-24.pdf","https://www.police1.com/police-robots/lapd-robot-dog-brings-peaceful-end-to-standoff-after-sniffing-out-armed-man-on-bus","https://www.nbclosangeles.com/investigations/lapd-reports-2-more-uses-of-its-robotic-dog/3359391/","https://www.ibtimes.com/lapd-robot-dog-spot-3745457","https://cityclerk.lacity.org/lacityclerkconnect/index.cfm?fa=ccfi.viewrecord&cfnumber=22-1438","https://oecd.ai/en/incidents/2024-10-02-1544","https://spectrumlocalnews.com/ca/california/public-safety/2025/01/01/police-robot-dog"],"source_refs":["AIBD-0379-source-0001","AIBD-0379-source-0002","AIBD-0379-source-0003","AIBD-0379-source-0004","AIBD-0379-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0379-reported-measure","evaluation_data":null},{"case_id":"AIBD-0183","case_url":"records/AIBD-0183.html","measure":"originated half of tax controls but accounted for 13.8% of the tax and penalties recovered in 2023","reported_result":"originated half of tax controls but accounted for 13.8% of the tax and penalties recovered in 2023","baseline":"le rendement moyen à l’initiative des agents était près de cinq fois supérieur","result":"celui des contrôles résultant du data mining","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.larevuedudigital.com/divergences-sur-lusage-du-data-mining-dans-le-controle-fiscal/","https://www.ccomptes.fr/fr/publications/la-detection-de-la-fraude-fiscale-des-particuliers","https://www.assemblee-nationale.fr/dyn/opendata/CRCANR5L17S2026PO59048N129.html","https://www.ccomptes.fr/fr/publications/la-lutte-contre-la-fraude-fiscale"],"source_refs":["AIBD-0183-source-0001","AIBD-0183-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"tax controls initiated by DGFiP agents' own analysis","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: le rendement moyen à l’initiative des agents était près de cinq fois supérieur. Result: celui des contrôles résultant du data mining.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Nearly a third of tax audits used mass-data risk analysis in the 2023 report","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0183-reported-measure","evaluation_data":null},{"case_id":"AIBD-0261","case_url":"records/AIBD-0261.html","measure":"2.2 million crystal structures (380,000 stable)","reported_result":"2.2 million crystal structures (380,000 stable)","baseline":"48,000 computationally stable materials","result":"a total of 421,000 stable crystals","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10700131/","https://deepmind.google/discover/blog/millions-of-new-materials-discovered-with-deep-learning/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11044265/","https://www.404media.co/google-says-it-discovered-millions-of-new-materials-with-ai-human-researchers/","https://cen.acs.org/research-integrity/Google-DeepMind-structures-near-duplicates-questionable/104/web/2026/10","https://journals.aps.org/prxenergy/abstract/10.1103/PRXEnergy.3.011002","https://pmc.ncbi.nlm.nih.gov/articles/PMC12307791/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12822528/","https://arxiv.org/abs/2610.13885","https://pmc.ncbi.nlm.nih.gov/articles/PMC12810037/"],"source_refs":["AIBD-0261-source-0001","AIBD-0261-source-0002","AIBD-0261-source-0003","AIBD-0261-source-0004","AIBD-0261-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"stable crystals previously identified by computational studies before GNoME","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 48,000 computationally stable materials. Result: a total of 421,000 stable crystals.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"2.2 million crystal structures (380,000 stable)","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0261-reported-measure","evaluation_data":null},{"case_id":"AIBD-0227","case_url":"records/AIBD-0227.html","measure":"charges were filed against one","reported_result":"charges were filed against one","baseline":"In the past, something like this would have been missed entirely.","result":"helped us stop this dangerous activity before anybody was hurt","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.govtech.com/public-safety/ai-gun-detection-system-leads-to-charges-in-new-mexico","https://lubbocklights.com/umc-hobbs-using-technology-to-spot-guns-prevent-tragedies/","https://zeroeyes.com/press-release/zeroeyes-positive-gun-detection-in-hobbs-new-mexico-leads-to-criminal-charges"],"source_refs":["AIBD-0227-source-0001","AIBD-0227-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: unmonitored security cameras before ZeroEyes, as described by Hobbs Police Chief August Fons","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: In the past, something like this would have been missed entirely. Result: helped us stop this dangerous activity before anybody was hurt.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"charges were filed against one","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0227-reported-measure","evaluation_data":null},{"case_id":"AIBD-0038","case_url":"records/AIBD-0038.html","measure":"confirmed Alström syndrome in two business days","reported_result":"confirmed Alström syndrome in two business days","baseline":"the genetic analysis of the sequenced exome can take up to months","result":"the suspected diagnosis of AS was confirmed within just two business days","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10783252/"],"source_refs":["AIBD-0038-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: untargeted analysis of the sequenced exome","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: the genetic analysis of the sequenced exome can take up to months. Result: the suspected diagnosis of AS was confirmed within just two business days.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Alström syndrome confirmed in two days","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0038-reported-measure","evaluation_data":null},{"case_id":"AIBD-0015","case_url":"records/AIBD-0015.html","measure":"cap set of size 512 in dimension 8, exceeding the previous best known construction of 496","reported_result":"cap set of size 512 in dimension 8, exceeding the previous best known construction of 496","baseline":"496","result":"512","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://cs.nyu.edu/~davise/papers/FunSearchComment.pdf","https://www.nature.com/articles/s41586-023-06924-6","https://www.technologyreview.com/2023/12/14/1085318/google-deepmind-large-language-model-solve-unsolvable-math-problem-cap-set/","https://arxiv.org/abs/2508.07932","https://arxiv.org/pdf/2408.02328","https://arxiv.org/pdf/2606.12194v2"],"source_refs":["AIBD-0015-source-0001","AIBD-0015-source-0002","AIBD-0015-source-0003","AIBD-0015-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the previous best known cap set construction in dimension 8","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 496. Result: 512.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"cap set of size 512 in dimension 8, exceeding the previous best known construction of 496","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0015-reported-measure","evaluation_data":null},{"case_id":"AIBD-0220","case_url":"records/AIBD-0220.html","measure":"Odds of completing each task relative to baseline; task time; System Usability Scale score.","reported_result":"With Envision Glasses, completion odds rose significantly for all six text tasks and for three of six searching and identifying tasks; street sign reading was significantly faster than baseline.","baseline":"Before, a bus would drive right by me because I'm standing at a yield sign","result":"Now he can ask the glasses for a description to ensure he's in the right place.","baseline_value":null,"result_value":null,"unit":null,"effect":"With Envision Glasses, completion odds rose significantly for all six text tasks and for three of six searching and identifying tasks; street sign reading was significantly faster than baseline.","design":"quasi-experimental.","sample":"25 adults with vision loss (acuity 20/100 to no light perception, mean age 64) from the Lighthouse Guild low-vision clinic, New York; no prior proficiency with the devices.","uncertainty":"Small sample in one clinic; single sessions after brief practice; participants were compensated; Envision's mean usability score was 62.5 and its median satisfaction 3 of 5; the glasses misread some text that participants corrected from context.","comparison_basis":"assumed","source_urls":["https://www.theglobeandmail.com/business/article-ai-smart-glasses-apps-vision/","https://www.letsenvision.com/glasses","https://pmc.ncbi.nlm.nih.gov/articles/PMC11721483/"],"source_refs":["AIBD-0220-source-0003"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The baseline had no AI device, so the measured gain is the uplift of the device itself.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The same participants without any of the tested devices, using their usual correction.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"With Envision Glasses, completion odds rose significantly for all six text tasks and for three of six searching and identifying tasks; street sign reading was significantly faster than baseline.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One session of about 90 minutes with each device.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"quasi-experimental.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Each participant did 14 timed and scored daily living tasks once without a device and once with each of four devices, in counterbalanced order over five clinic sessions; usability by the System Usability Scale.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Odds of completing each task relative to baseline; task time; System Usability Scale score.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"25 adults with vision loss (acuity 20/100 to no light perception, mean age 64) from the Lighthouse Guild low-vision clinic, New York; no prior proficiency with the devices.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small sample in one clinic; single sessions after brief practice; participants were compensated; Envision's mean usability score was 62.5 and its median satisfaction 3 of 5; the glasses misread some text that participants corrected from context.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0220-reported-measure","evaluation_data":null},{"case_id":"AIBD-0036","case_url":"records/AIBD-0036.html","measure":"route around the area or transit at 10 knots or less","reported_result":"route around the area or transit at 10 knots or less","baseline":null,"result":"more slowdowns than aerial surveillance by a factor of two to five during fall and summer","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://tos.org/oceanography/article/glider-surveillance-for-near-real-time-detection-and-spatial-management-of-north-atlantic-right-whales","https://content.govdelivery.com/accounts/USNOAAFISHERIES/bulletins/3824b02","https://dcs.whoi.edu/lfdcs_manual/lfdcs_manual.html","https://repository.library.noaa.gov/view/noaa/48671","https://www.fisheries.noaa.gov/national/endangered-species-conservation/reducing-vessel-strikes-north-atlantic-right-whales","https://oceana.org/press-releases/oceana-finds-most-boats-speeding-in-slow-zones-designed-to-protect-critically-endangered-north-atlantic-right-whales/"],"source_refs":["AIBD-0036-source-0001","AIBD-0036-source-0002","AIBD-0036-source-0003","AIBD-0036-source-0004","AIBD-0036-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"aerial surveillance","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Result: more slowdowns than aerial surveillance by a factor of two to five during fall and summer.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"mariners route around the area or transit at 10 knots or less","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0036-reported-measure","evaluation_data":null},{"case_id":"AIBD-0104","case_url":"records/AIBD-0104.html","measure":"Average vehicle speed change when the warning showed; detection precision and recall.","reported_result":"drivers slowed by an average of 6.30 km/h and 5.06 km/h at two trial zones","baseline":"Speeds without a warning on the sign (absolute values not given in the summary)","result":"Average speeds 6.30 km/h and 5.06 km/h lower at the two event-zone sites when a warning showed","baseline_value":null,"result_value":null,"unit":null,"effect":"Speeds fell by 6.30 km/h and 5.06 km/h at the two event-zone sites and by 4.26 km/h and 3.44 km/h at the two approach-zone sites; precision 0.77 and recall 0.97.","design":"other.","sample":"One cassowary crossing on the Kennedy Highway at Kuranda, Far North Queensland; five-month trial from late January 2024 with two months of on-road warnings; 287 manually verified cassowary sightings.","uncertainty":"The summary gives no confidence intervals, counts of vehicles or absolute speeds. Crash reductions are modeled from speeds and were not observed. The posted limit at the site was 60 km/h during roadwork. The developers wrote the evaluation.","comparison_basis":"measured","source_urls":["https://imoveaustralia.com/wp-content/uploads/2025/04/Summary-The-Development-and-Performance-Testing-of-a-LAARMA%E2%80%94Large-Animal-Activ.pdf","https://www.abc.net.au/news/2025-07-30/ai-saves-cassowary-lives-reducing-vehicle-strikes-far-north-qld/105589446","https://imoveaustralia.com/project/large-australian-animals-and-cars-safety-for-all/","https://techxplore.com/news/2025-08-world-ai-roadside-technology-animalvehicle.html","https://news.mongabay.com/2026/05/amid-efforts-to-save-australias-southern-cassowaries-their-numbers-remain-unknown/"],"source_refs":["AIBD-0104-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The detector triggered every warning, so the speed changes measure the effect of the AI-triggered alerts.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Speeds at the same sites without a warning on the sign.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Speeds fell by 6.30 km/h and 5.06 km/h at the two event-zone sites and by 4.26 km/h and 3.44 km/h at the two approach-zone sites; precision 0.77 and recall 0.97.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Drivers passing one site during two months of on-road warnings.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Speed sensors at four sites (two in the approach zone, two in the event zone) recorded vehicle speeds while the sign showed or did not show a warning; detections were checked against manually verified cassowary sightings; crash effects were modeled with the Nilsson power model.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Average vehicle speed change when the warning showed; detection precision and recall.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The developers state that the source code is freely available in two GitHub repositories, acfr/CassDetect and acfr/laarma.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One cassowary crossing on the Kennedy Highway at Kuranda, Far North Queensland; five-month trial from late January 2024 with two months of on-road warnings; 287 manually verified cassowary sightings.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The summary gives no confidence intervals, counts of vehicles or absolute speeds. Crash reductions are modeled from speeds and were not observed. The posted limit at the site was 60 km/h during roadwork. The developers wrote the evaluation.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0104-reported-measure","evaluation_data":null},{"case_id":"AIBD-0127","case_url":"records/AIBD-0127.html","measure":"he responded within one or two weeks","reported_result":"he responded within one or two weeks","baseline":"Coates was now entering a terminal stage: his heart and kidneys were failing, and fluid had to be drained from his belly several times a week. He was too sick to undergo a stem-cell transplant, which could save his life.","result":"within days Coates started to improve. A few months later, he'd recovered sufficiently to receive a stem-cell transplant. He is now in remission.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.newyorker.com/culture/open-questions/can-ai-find-cures-for-untreatable-diseases-using-drugs-we-already-have","https://www.nytimes.com/2025/03/20/well/ai-drug-repurposing.html","https://everycure.org/portfolio/","https://everycure.org/wp-content/uploads/2026/01/EVery.Cure_.2025.Annual.Report.Single.Digital.Pages.pdf","https://arpa-h.gov/sites/default/files/2026-04/FY27%20ARPA-H%20CJ_Greenlight_508%20compliant.pdf","https://www.youtube.com/watch?v=4FxwoOuPGWs","https://www.aamc.org/news/doctor-saved-his-own-life-now-he-s-mission-save-thousands-more","https://doi.org/10.1016/S2352-3026(24)00278-3","https://whyy.org/articles/ai-platform-every-cure-medicine-disease-treatment-university-pennyslvania/","https://www.elsevier.com/en-gb/connect/elsevier-and-every-cure-treating-rare-diseases-with-generic-drugs-at-scale","https://pmc.ncbi.nlm.nih.gov/articles/PMC10229532/"],"source_refs":["AIBD-0127-source-0001","AIBD-0127-source-0002","AIBD-0127-source-0003","AIBD-0127-source-0004","AIBD-0127-source-0005","AIBD-0127-source-0006","AIBD-0127-source-0007","AIBD-0127-source-0008","AIBD-0127-source-0009","AIBD-0127-source-0010"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the treatments his doctors had tried, with hospice as the remaining option","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: Coates was now entering a terminal stage: his heart and kidneys were failing, and fluid had to be drained from his belly several times a week. He was too sick to undergo a stem-cell transplant, which could save his life. Result: within days Coates started to improve. A few months later, he'd recovered sufficiently to receive a stem-cell transplant. He is now in remission.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"he responded within one or two weeks","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0127-reported-measure","evaluation_data":null},{"case_id":"AIBD-0263","case_url":"records/AIBD-0263.html","measure":"about 10 hours of daily use for chess, coursework, scheduling and communication","reported_result":"about 10 hours of daily use for chess, coursework, scheduling and communication","baseline":"not something I was able to do beforehand","result":"I can control a computer just like anyone else can","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://abcnews.com/GMA/Wellness/neuralinks-brain-implant-patient-feared-device-removed/story?id=110325322","https://www.abc15.com/news/business/elon-musks-neuralink-begins-clinical-trials-in-phoenix","https://www.barrowneuro.org/about/news-and-articles/patient-stories/barrow-patient-begins-to-reclaim-independence-with-neuralink-robotic-arm-after-spinal-cord-injury/","https://www.techtimes.com/articles/317356/20260528/neuralink-bci-after-28-months-arbaugh-tells-robotics-summit-what-published-research-cannot.htm","https://thedebrief.org/neuralink-hits-7-users-as-global-brain-implant-count-remains-under-100/","https://www.nature.com/articles/d41586-024-02368-8","https://web.archive.org/web/2026/https://news.sky.com/story/man-given-musks-neuralink-brain-chip-in-uk-trial-says-it-feels-magical-and-gives-new-hope-13500126"],"source_refs":["AIBD-0263-source-0001","AIBD-0263-source-0002","AIBD-0263-source-0003","AIBD-0263-source-0004","AIBD-0263-source-0005","AIBD-0263-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: Arbaugh's own ability to use a computer before the implant (self-report)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: not something I was able to do beforehand. Result: I can control a computer just like anyone else can.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"10 hours of daily use for chess, coursework, scheduling and communication at 28 months","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0263-reported-measure","evaluation_data":null},{"case_id":"AIBD-0393","case_url":"records/AIBD-0393.html","measure":"Participant-reported experience and satisfaction.","reported_result":"Participants reported physical, psychological and social satisfaction and a wish to recommend the service.","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":"Participants reported physical, psychological and social satisfaction and a wish to recommend the service.","design":"qualitative evaluation.","sample":"40 older adults across 10 smart senior centers in Gimhae, South Korea, using the service for six weeks or more.","uncertainty":"Self-reported outcomes, no control group and a small sample. The research team worked under an industry-academia agreement with ESTsoft, and ESTsoft lists Yonsei University as a co-developer of the exercise modules. The paper (Korean Journal of Physical Education, volume 63, issue 3) is known only through the developer's release.","comparison_basis":null,"source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://estsoft.ai/en/all/240708"],"source_refs":["AIBD-0393-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated; Professor Huh Jin-moo of Yonsei University says that the content shows effects similar to conventional leisure activity.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"None reported.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"qualitative evaluation.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Participants reported physical, psychological and social satisfaction and a wish to recommend the service.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Six weeks or more of use in the centers.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"qualitative evaluation.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"In-depth interviews analyzed in three stages (introduction, adaptation, acceptance) through diffusion of innovation theory.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Participant-reported experience and satisfaction.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"40 older adults across 10 smart senior centers in Gimhae, South Korea, using the service for six weeks or more.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Self-reported outcomes, no control group and a small sample. The research team worked under an industry-academia agreement with ESTsoft, and ESTsoft lists Yonsei University as a co-developer of the exercise modules. The paper (Korean Journal of Physical Education, volume 63, issue 3) is known only through the developer's release.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0393-reported-measure","evaluation_data":null},{"case_id":"AIBD-0426","case_url":"records/AIBD-0426.html","measure":"reduced assessment penalties by 27% and 15%","reported_result":"reduced assessment penalties by 27% and 15%","baseline":"not stated","result":"penalties lower by 27% and 15% in the first half of 2024","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-260","https://web.archive.org/web/2026/https://e.huawei.com/cn/news/2025/industries/grid/huadian-huawei-ai-for-good"],"source_refs":["AIBD-0426-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0426-reported-measure","evaluation_data":null},{"case_id":"AIBD-0447","case_url":"records/AIBD-0447.html","measure":"widened the monitored area of a Xiamen reserve from 120 to 330 square kilometers","reported_result":"widened the monitored area of a Xiamen reserve from 120 to 330 square kilometers","baseline":"120 square kilometers monitored","result":"330 square kilometers monitored","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-387","https://www.eqs-news.com/news/corporate-news/ai-solution-shows-early-success-in-protecting-chinese-white-dolphins/42998d9d-9298-4044-b587-bf438525c563_en","https://web.archive.org/web/2026/https://www.tmforum.org/catalysts/projects/C24.5.747/wildlife-guardian-ai-5g-advanced-sustainable-application"],"source_refs":["AIBD-0447-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0447-reported-measure","evaluation_data":null},{"case_id":"AIBD-0484","case_url":"records/AIBD-0484.html","measure":"Control error in plasma shape, radial position and vertical X-point position.","reported_result":"Shape errors within 1.5 to 2.0 cm and radial errors within 1 cm; X-point errors of up to about 4 cm.","baseline":"shape control computed errors from a real-time reconstruction and fed them to PID controllers","result":"the RL controller mapped raw magnetic signals to coil commands; shape errors stayed within 1.5 to 2.0 cm and radial errors within 1 cm, with X-point errors of up to about 4 cm","baseline_value":null,"result_value":null,"unit":null,"effect":"Shape errors within 1.5 to 2.0 cm and radial errors within 1 cm; X-point errors of up to about 4 cm.","design":"other.","sample":"Four controllers tested at DIII-D across heating power scans in L-mode and H-mode plasmas, including L-H transitions and pellet injections.","uncertainty":"All authors work for the developer or the facility or receive support from the developer. The X-point mismatch shows limits of the simulator used for training.","comparison_basis":"assumed","source_urls":["https://iopscience.iop.org/article/10.1088/1741-4326/ae34c6","https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0484-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Target shape parameters; no head-to-head comparison with isoflux control is reported.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Shape errors within 1.5 to 2.0 cm and radial errors within 1 cm; X-point errors of up to about 4 cm.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Experimental discharges at DIII-D during 2023 and 2024 by the developer's account.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Experimental discharges on the DIII-D tokamak with the RL controller in the Plasma Control System at 4 kHz; errors computed against target shape parameters.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Control error in plasma shape, radial position and vertical X-point position.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Four controllers tested at DIII-D across heating power scans in L-mode and H-mode plasmas, including L-H transitions and pellet injections.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"All authors work for the developer or the facility or receive support from the developer. The X-point mismatch shows limits of the simulator used for training.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0484-reported-measure","evaluation_data":null},{"case_id":"AIBD-0516","case_url":"records/AIBD-0516.html","measure":null,"reported_result":null,"baseline":"12 months","result":"15 days","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://govmedia.com/co-written-partner/event-news/mohammed-bin-rashid-housing-establishment-earns-two-wins-govmedia-conference-awards-2025","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-811","https://mediaoffice.ae/en/news/2023/october/05-10/the-smart-housing-matrix-of-mohammed-bin-rashid-housing-establishment-among-global-best-practices"],"source_refs":["AIBD-0516-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0516-reported-measure","evaluation_data":null},{"case_id":"AIBD-0523","case_url":"records/AIBD-0523.html","measure":null,"reported_result":null,"baseline":"About three hours of recording work per session","result":"About five minutes per session","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-854","https://news.sktelecom.com/226306","https://dealsite.co.kr/articles/170093","https://bizhankook.com/articles/29204.html"],"source_refs":["AIBD-0523-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0523-reported-measure","evaluation_data":null},{"case_id":"AIBD-0542","case_url":"records/AIBD-0542.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://web.archive.org/web/20261017031527/https://www.local.gov.uk/case-studies/barnsley-council-pioneering-use-microsoft-copilot","https://www.local.gov.uk/case-studies/barnsley-council-pioneering-use-microsoft-copilot","https://www.bbc.co.uk/news/articles/cj0n6q55gz4o"],"source_refs":["AIBD-0542-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0542-reported-measure","evaluation_data":null},{"case_id":"AIBD-0561","case_url":"records/AIBD-0561.html","measure":null,"reported_result":null,"baseline":"The previous year's call volumes, not stated","result":"A noticeable reduction in call volumes, not quantified","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/south-cambridgeshire-contact-centre-chatbot-improve-resident-satisfaction","https://www.local.gov.uk/case-studies/south-cambridgeshire-contact-centre-chatbot-improve-resident-satisfaction","https://www.cambridge-news.co.uk/news/cambridge-news/four-day-week-council-says-34084986"],"source_refs":["AIBD-0561-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0561-reported-measure","evaluation_data":null},{"case_id":"AIBD-0565","case_url":"records/AIBD-0565.html","measure":"Number of detected near misses; average vehicle speed at recorded near misses.","reported_result":"17 near misses before and 2 after (more than 88 percent fewer); speeds at near misses fell from 14 to 16 mph to 10.5 to 12.5 mph.","baseline":"17 near misses in two months","result":"2 near misses in the same length of period","baseline_value":null,"result_value":null,"unit":null,"effect":"17 near misses before and 2 after (more than 88 percent fewer); speeds at near misses fell from 14 to 16 mph to 10.5 to 12.5 mph.","design":"before/after.","sample":"One T-junction, Grange Road and Anderton Road, Longford, Coventry.","uncertainty":"Small counts over short windows at a single site, chosen because near misses were high, so regression to the mean is possible. No control site, no collision or injury data, and the figures come from TfWM.","comparison_basis":"measured","source_urls":["https://www.bbc.co.uk/news/articles/c8r0703kv18o","https://www.local.gov.uk/case-studies/transport-west-midlands-near-miss-detection-vivacity","https://www.wmca.org.uk/news/junction-sees-huge-cut-in-near-misses-thanks-to-ai-powered-safety-technology/","https://www.wmca.org.uk/news/new-ai-technology-reveals-road-safety-risks-across-the-west-midlands/","https://www.traffictechnologytoday.com/features/feature-how-near-miss-data-is-revolutionizing-road-safety.html"],"source_refs":["https://www.bbc.co.uk/news/articles/c8r0703kv18o"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated; the change follows a physical road measure that the sensor data helped to select.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The same junction before the refuge.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"17 near misses before and 2 after (more than 88 percent fewer); speeds at near misses fell from 14 to 16 mph to 10.5 to 12.5 mph.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two months before and two months after the refuge was built.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Near misses counted by the VivaCity sensor at the junction over two-month windows before and after the refuge was built.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Number of detected near misses; average vehicle speed at recorded near misses.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One T-junction, Grange Road and Anderton Road, Longford, Coventry.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small counts over short windows at a single site, chosen because near misses were high, so regression to the mean is possible. No control site, no collision or injury data, and the figures come from TfWM.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0565-reported-measure","evaluation_data":null},{"case_id":"AIBD-0592","case_url":"records/AIBD-0592.html","measure":"cut cost per impression by 9% and cost per click by 5%","reported_result":"cut cost per impression by 9% and cost per click by 5%","baseline":"Click-through rate of human-only and AI-only copy","result":"Combined human and AI copy had a click-through rate 3% higher than either","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/20251009235439/https://www.battlegroundai.com/resources/from-research-to-results-how-2024-became-a-year-of-innovation-and-opportunity","https://www.battlegroundai.com/","https://docs.google.com/spreadsheets/d/1ZDVVXiGeMcKJlq4F3GvvZALDtLliGeJtdDV6Zhr_p-0/edit#gid=0"],"source_refs":["AIBD-0592-source-0001","AIBD-0592-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0592-reported-measure","evaluation_data":null},{"case_id":"AIBD-0616","case_url":"records/AIBD-0616.html","measure":"Minutes per week of lesson and resource preparation.","reported_result":"saving of 25.3 minutes a week, or 31 percent","baseline":"81.5 minutes a week","result":"About 56.2 minutes a week","baseline_value":null,"result_value":null,"unit":null,"effect":"56.2 against 81.5 minutes a week (69 percent of the comparison group's time).","design":"randomized.","sample":"259 Year 7 and Year 8 science teachers in 68 schools, ten weeks in the 2024 summer term.","uncertainty":"Time was self-reported; teachers used the approach for part of their timetable; student outcomes were not measured.","comparison_basis":"measured","source_urls":["https://www.nfer.ac.uk/publications/chatgpt-in-lesson-preparation-a-teacher-choices-trial/"],"source_refs":["AIBD-0616-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The time saving is attributed to ChatGPT use with the guide.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Teachers asked not to use generative AI.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"56.2 against 81.5 minutes a week (69 percent of the comparison group's time).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Ten weeks of lesson preparation for Year 7 and Year 8 science.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Teacher-reported weekly lesson and resource preparation time; blind expert panel review of sampled resources.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Minutes per week of lesson and resource preparation.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"259 Year 7 and Year 8 science teachers in 68 schools, ten weeks in the 2024 summer term.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time was self-reported; teachers used the approach for part of their timetable; student outcomes were not measured.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0616-reported-measure","evaluation_data":null},{"case_id":"AIBD-0077","case_url":"records/AIBD-0077.html","measure":"a maternally inherited deep intronic PKHD1 variant scoring 0.98","reported_result":"a maternally inherited deep intronic PKHD1 variant scoring 0.98","baseline":"Only the paternal pathogenic missense variant was found; no maternally inherited coding variant explained the disease.","result":"SpliceAI nominated a maternal deep intronic variant with a score of 0.98, which RNA studies confirmed as the second allele.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11116050/"],"source_refs":["AIBD-0077-source-0001","AIBD-0077-source-0002","AIBD-0077-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a maternally inherited deep intronic PKHD1 variant scoring 0.98","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a maternally inherited deep intronic PKHD1 variant scoring 0.98","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0077-reported-measure","evaluation_data":null},{"case_id":"AIBD-0630","case_url":"records/AIBD-0630.html","measure":"Study 1: Referral sensitivity and specificity. Study 2: Sensitivity and specificity.","reported_result":"Their sensitivity rose from 82.0% to 91.4%","baseline":"Sensitivity 82.0%, specificity 44.2%","result":"Sensitivity 91.4%, specificity 32.4%","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: Sensitivity 82.0% to 91.4%; specificity 44.2% to 32.4%. Study 2: Sensitivity 95.5% against 83.0%; specificity 20.7% against 54.2%.","design":"Study 1: quasi-experimental. Study 2: other.","sample":"Study 1: 108 internal medicine or family practice physicians; 25 malignant and 25 benign lesions. Study 2: 1,005 patients, 1,579 lesions, 22 sites in the US and Australia.","uncertainty":"Study 1: Image-based reads; lower specificity; limited data for Fitzpatrick IV to VI. Study 2: Low specificity; under-40 sensitivity below 90%.","comparison_basis":"measured","source_urls":["https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN230008.pdf"],"source_refs":["AIBD-0630-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: AUROC 0.708 to 0.762. Study 2: Not a with-and-without comparison.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Unaided reads by the same physicians. Study 2: Primary care physicians on the same lesions.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: quasi-experimental. Study 2: other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Sensitivity 82.0% to 91.4%; specificity 44.2% to 32.4%. Study 2: Sensitivity 95.5% against 83.0%; specificity 20.7% against 54.2%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Two reads per lesion. Study 2: One device reading per lesion.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: quasi-experimental. Study 2: other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Multi-reader study; each lesion image viewed with and without device output in random order; refer or do not refer decision. Study 2: Prospective blinded standalone accuracy study against biopsy, with physician assessment on the same lesions.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Referral sensitivity and specificity. Study 2: Sensitivity and specificity.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: None public. Study 2: None public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 108 internal medicine or family practice physicians; 25 malignant and 25 benign lesions. Study 2: 1,005 patients, 1,579 lesions, 22 sites in the US and Australia.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Image-based reads; lower specificity; limited data for Fitzpatrick IV to VI. Study 2: Low specificity; under-40 sensitivity below 90%.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0630-reported-measure","evaluation_data":null},{"case_id":"AIBD-0055","case_url":"records/AIBD-0055.html","measure":"Active range of motion of the thumb and wrist; pain levels.","reported_result":"active thumb extension (4 degrees) and active wrist flexion (6 degrees)","baseline":"no active movement prior the MME interventions","result":"active extension of the thumb (4 degrees) and regained active wrist movement (flex: 6 degrees, extend: 10 degrees)","baseline_value":null,"result_value":null,"unit":null,"effect":"Active thumb extension of 4 degrees, wrist flexion of 6 degrees and wrist extension of 10 degrees; no consistent change in pain.","design":"case study.","sample":"One 72-year-old man with a replanted right arm and no functional recovery after conventional rehabilitation, Sweden.","uncertainty":"A single uncontrolled case; no follow-up after the final session; two authors hold a pending patent on the technology.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11585237/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11742229/"],"source_refs":["AIBD-0055-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated; the patient also received sensory training in nine sessions, and no session without the decoder was run.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The same arm before treatment, which had no active thumb extension or wrist movement.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Active thumb extension of 4 degrees, wrist flexion of 6 degrees and wrist extension of 10 degrees; no consistent change in pain.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"18 treatment sessions of about two hours, nine with sensory training, over 84 days.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Goniometer range of motion at the first and final assessment sessions and at session 10; weighted pain distribution and numeric pain rating; Semmes-Weinstein monofilament test used in an exploratory way.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Active range of motion of the thumb and wrist; pain levels.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One 72-year-old man with a replanted right arm and no functional recovery after conventional rehabilitation, Sweden.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"A single uncontrolled case; no follow-up after the final session; two authors hold a pending patent on the technology.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0055-reported-measure","evaluation_data":null},{"case_id":"AIBD-0229","case_url":"records/AIBD-0229.html","measure":"the result came back well north of 99 percent for ElevenLabs","reported_result":"the result came back well north of 99 percent for ElevenLabs","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.wired.com/story/biden-robocall-deepfake-elevenlabs/","https://www.nhpr.org/politics/2024-05-23/political-operative-linked-to-ai-biden-robocalls-facing-charges-in-nh-fcc-fines","https://www.pindrop.com/resources/article/pindrop-reveals-tts-engine-behind-biden-ai-robocall","https://docs.fcc.gov/public/attachments/FCC-24-59A1.pdf"],"source_refs":["AIBD-0229-source-0001","AIBD-0229-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the result came back well north of 99 percent for ElevenLabs","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the result came back well north of 99 percent for ElevenLabs","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0229-reported-measure","evaluation_data":null},{"case_id":"AIBD-0123","case_url":"records/AIBD-0123.html","measure":"Grasp-detection F1-score; ARAT score.","reported_result":"drink water unaided, with grasp-decoding accuracy above 90%","baseline":"处于四肢瘫痪状态已经14年 (no hand function for 14 years)","result":"实现自主喝水等脑控功能，抓握解码准确率超过90% (brain-controlled drinking, grasp decoding accuracy above 90%)","baseline_value":null,"result_value":null,"unit":null,"effect":"Average F1-score of 0.91, 100% success in object transfer tests, a 5-point gain in upper limb motor score and a 27-point gain in ARAT.","design":"case study.","sample":"One man with complete cervical spinal cord injury, implanted at Xuanwu Hospital and trained at home.","uncertainty":"A single patient; the authors are the system's developers and clinicians; the preprint has not been peer reviewed.","comparison_basis":"measured","source_urls":["https://news.sciencenet.cn/htmlnews/2024/1/516951.shtm","https://doi.org/10.1101/2024.09.05.24313041","https://news.cctv.com/2024/01/31/ARTIAm1lfoEPddPWL8sZzJPd240131.shtml","https://news.bjd.com.cn/2024/01/31/10691652.shtml","https://clinicaltrials.gov/study/NCT05920174","https://www.tsinghua.edu.cn/info/1182/109663.htm","https://www.scientificamerican.com/article/china-just-approved-its-first-brain-implant-for-commercial-use-a-world-first/"],"source_refs":["AIBD-0123-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The decoder detects grasp intention and triggers the glove; the study does not separate the decoder's share of the neurological recovery from that of repeated training.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The patient's own scores before BCI training.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Average F1-score of 0.91, 100% success in object transfer tests, a 5-point gain in upper limb motor score and a 27-point gain in ARAT.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Nine months of home use.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Grasp-detection F1-score, object transfer tests, upper limb motor score, Action Research Arm Test and somatosensory evoked potentials over nine months of home use.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Grasp-detection F1-score; ARAT score.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One man with complete cervical spinal cord injury, implanted at Xuanwu Hospital and trained at home.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"A single patient; the authors are the system's developers and clinicians; the preprint has not been peer reviewed.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0123-reported-measure","evaluation_data":null},{"case_id":"AIBD-0301","case_url":"records/AIBD-0301.html","measure":"he received four bypass grafts; recovery was uneventful","reported_result":"he received four bypass grafts; recovery was uneventful","baseline":"A standard ECG was completely normal, as was physical examination.","result":"he noticed immediately after exercise, a negative T-wave that became more prominent in ∼7 min, then slowly returned back to a normal positive T-wave after 22 min","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC10872691/","https://pmc.ncbi.nlm.nih.gov/articles/PMC7466842/"],"source_refs":["AIBD-0301-source-0001","AIBD-0301-source-0002","AIBD-0301-source-0003","AIBD-0301-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the cardiologist's resting 12-lead ECG and physical examination","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: A standard ECG was completely normal, as was physical examination. Result: he noticed immediately after exercise, a negative T-wave that became more prominent in ∼7 min, then slowly returned back to a normal positive T-wave after 22 min.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"he received four bypass grafts; recovery was uneventful","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0301-reported-measure","evaluation_data":null},{"case_id":"AIBD-0551","case_url":"records/AIBD-0551.html","measure":"Time to complete case notes, assessments and supervision write-ups.","reported_result":"Average time cut of 50% to 60% for notes and assessments; supervision write-ups fell from 40 minutes to under 10.","baseline":"About 40 minutes to write up a supervision","result":"Under 10 minutes to write up a supervision","baseline_value":null,"result_value":null,"unit":null,"effect":"Average time cut of 50% to 60% for notes and assessments; supervision write-ups fell from 40 minutes to under 10.","design":"before/after.","sample":"Kingston adult social care pilot that started with three social workers and grew to more than 10.","uncertainty":"Self-reported times with no control group and no published sample counts; the council reports poor transcripts for people with low verbal skills or limited capacity and for poor audio.","comparison_basis":"measured","source_urls":["http://web.archive.org/web/20261021233345/https://www.local.gov.uk/case-studies/kingston-council-using-ai-adult-social-care-administration","https://www.local.gov.uk/case-studies/kingston-council-using-ai-adult-social-care-administration","https://diginomica.com/how-two-uk-councils-are-using-ai-note-taking-cut-social-workers-admin-burden","https://www.theguardian.com/society/2024/sep/28/social-workers-england-ai-system-magic-notes","https://www.adalovelaceinstitute.org/report/scribe-and-prejudice/"],"source_refs":["http://web.archive.org/web/20261021233345/https://www.local.gov.uk/case-studies/kingston-council-using-ai-adult-social-care-administration"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The same staff's time to write records without the tool.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Average time cut of 50% to 60% for notes and assessments; supervision write-ups fell from 40 minutes to under 10.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Staff feedback on time saved, gathered by Beam and the council in a weekly digest.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time to complete case notes, assessments and supervision write-ups.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Kingston adult social care pilot that started with three social workers and grew to more than 10.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Self-reported times with no control group and no published sample counts; the council reports poor transcripts for people with low verbal skills or limited capacity and for poor audio.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0551-reported-measure","evaluation_data":null},{"case_id":"AIBD-0560","case_url":"records/AIBD-0560.html","measure":null,"reported_result":null,"baseline":"72 minutes to draft meeting minutes and actions","result":"17 minutes with Copilot","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/somerset-council-using-microsoft-copilot-improve-productivity","https://www.local.gov.uk/case-studies/somerset-council-using-microsoft-copilot-improve-productivity","https://www.microsoft.com/en/customers/story/22699-somerset-council-microsoft-365-copilot"],"source_refs":["AIBD-0560-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0560-reported-measure","evaluation_data":null},{"case_id":"AIBD-0587","case_url":"records/AIBD-0587.html","measure":"Error in estimated ore volume and remaining uncertainty after each measurement.","reported_result":"appraised the high-grade zone in about eight months","baseline":"Exploration drilling on a regular grid, guided by standard geostatistics.","result":"Drill holes chosen in sequence to cut uncertainty on grade, thickness and geology; the high-grade zone was appraised in about eight months.","baseline_value":null,"result_value":null,"unit":null,"effect":"The authors report that the sequential planner reduced uncertainty more than the conventional patterns.","design":"simulation.","sample":"Synthetic two-dimensional exploration problems with one or two ore bodies; no field data.","uncertainty":"One author works for KoBold. The tests are synthetic and do not show the size of any gain at Mingomba.","comparison_basis":"assumed","source_urls":["https://www.mongolianminingjournal.com/a/73483","https://www.pwc.com/zm/en/assets/pdf/zambia-mining-report-2024.pdf","https://iea.blob.core.windows.net/assets/de9dea13-b07d-42c5-a398-d1b3ae17d866/EnergyandAI.pdf","https://koboldmetals.com/mingomba/","https://gmd.copernicus.org/articles/16/289/2023/","https://africanminingmarket.com/shaft-sinking-project-at-mingomba-mine-kicks-off/25540/","https://africanminingmarket.com/kobold-sets-early-2030s-first-copper-production-at-mingomba-project/25142/"],"source_refs":["AIBD-0587-source-0005"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Grid patterns of 1, 4, 9 and 16 measurements and random measurement locations.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"simulation.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"The authors report that the sequential planner reduced uncertainty more than the conventional patterns.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"simulation.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Monte Carlo trials of the sequential planner and of baseline drilling patterns on synthetic ore bodies.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Error in estimated ore volume and remaining uncertainty after each measurement.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The authors deposited code on Zenodo (doi 10.5281/zenodo.6727378).","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Synthetic two-dimensional exploration problems with one or two ore bodies; no field data.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One author works for KoBold. The tests are synthetic and do not show the size of any gain at Mingomba.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0587-reported-measure","evaluation_data":null},{"case_id":"AIBD-0582","case_url":"records/AIBD-0582.html","measure":null,"reported_result":null,"baseline":"the bleeding had not yet been detected by emergency room doctors or radiologists","result":"the AI picked up the bleeding before emergency room doctors or radiologists had detected it","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://cbs12.com/news/local/from-caregiver-to-cared-for-how-an-icu-nurses-training-saved-her-own-life","https://baptisthealth.net/baptist-health-news/stroke-nurse-shares-her-experience-after-brain-aneurysm","https://www.neurologylive.com/view/incorporating-ai-into-stroke-triage","https://thecoastalstar.com/profiles/blogs/fast-action-expert-surgeon-empowers-nurse-to-recover","https://www.viz.ai/trust-center-privacy-notice"],"source_refs":["AIBD-0582-source-0001","AIBD-0582-source-0002","AIBD-0582-source-0003","AIBD-0582-source-0004","AIBD-0582-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0582-reported-measure","evaluation_data":null},{"case_id":"AIBD-0013","case_url":"records/AIBD-0013.html","measure":"Study 1: Four passages of 140 characters each, with at least 85 percent of characters recoverable. Study 2: Agreement between model-recovered ink and ink visible in the new scan.","reported_result":"recovered 15 columns and more than 2,000 characters of Greek text","baseline":"0% of a scroll","result":"5% of a scroll; 15 columns of never-before-seen text, more than 2000 characters total","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: The winning entry met the criteria and added 11 more columns, for more than 2,000 characters in total. Study 2: The authors state that the new scan confirms the 2023 reading.","design":"Study 1: qualitative evaluation. Study 2: other.","sample":"Study 1: Grand Prize submissions received by 1 January 2024 for the scroll PHerc. Paris. 4. Study 2: PHerc. Paris. 4, Institut de France.","uncertainty":"Study 1: The review team was chosen by the organizers. The scroll was about 5 percent read. Study 2: A preprint without peer review, written by the organizers and one of the 2023 prize winners.","comparison_basis":"measured","source_urls":["https://scrollprize.org/grandprize","https://scrollprize.org/firstletters","https://uknow.uky.edu/research/grand-prize-discovery-made-2000-year-old-herculaneum-scrolls","https://www.nature.com/articles/d41586-024-00346-8","https://www.bodleian.ox.ac.uk/about/media/feb25/herculaneum-scroll","https://arxiv.org/html/2606.29085v1"],"source_refs":["AIBD-0013-source-0001","AIBD-0013-source-0006"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: The criteria that the organizers set when they launched the contest in March 2023.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: qualitative evaluation. Study 2: other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: The winning entry met the criteria and added 11 more columns, for more than 2,000 characters in total. Study 2: The authors state that the new scan confirms the 2023 reading.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: qualitative evaluation. Study 2: other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Papyrologists read anonymized submissions independently of one another, and a technical team reproduced the submitted code and methods. Study 2: The team rescanned PHerc. Paris. 4 at 2.4 micrometers on the BM18 beamline, segmented directly visible ink in three dimensions and compared it with the output of the 2023 prize-winning model.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Four passages of 140 characters each, with at least 85 percent of characters recoverable. Study 2: Agreement between model-recovered ink and ink visible in the new scan.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Grand Prize submissions received by 1 January 2024 for the scroll PHerc. Paris. 4. Study 2: PHerc. Paris. 4, Institut de France.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: The review team was chosen by the organizers. The scroll was about 5 percent read. Study 2: A preprint without peer review, written by the organizers and one of the 2023 prize winners.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0013-reported-measure","evaluation_data":null},{"case_id":"AIBD-0656","case_url":"records/AIBD-0656.html","measure":"Change in referral volume, overall and by group.","reported_result":"referrals rise by 15%, against 6% in control services","baseline":"6% rise in referrals (control services)","result":"15% rise in referrals (services using Limbic Access)","baseline_value":null,"result_value":null,"unit":null,"effect":"15% against 6% rise; nonbinary 179%; ethnic minority 29%.","design":"controlled observational.","sample":"129,400 patients in NHS Talking Therapies services in England.","uncertainty":"Observational; company authors; NICE notes mostly company-sourced evidence.","comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/38317020/","https://www.nice.org.uk/guidance/htg756/chapter/3-Committee-discussion","https://www.nice.org.uk/guidance/htg756/chapter/1-Recommendations"],"source_refs":["AIBD-0656-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"About 9 percentage points more growth in referrals.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Control services without the chatbot.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"15% against 6% rise; nonbinary 179%; ethnic minority 29%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Self-referral through the chatbot.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Referral volumes and demographics in routine service data, services with the chatbot against control services.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Change in referral volume, overall and by group.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"None public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"129,400 patients in NHS Talking Therapies services in England.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Observational; company authors; NICE notes mostly company-sourced evidence.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0656-reported-measure","evaluation_data":null},{"case_id":"AIBD-0101","case_url":"records/AIBD-0101.html","measure":"about 9,481 safe elephant crossings","reported_result":"about 9,481 safe elephant crossings","baseline":"Eleven wild elephants were killed on railway tracks in Madukkarai stretch from 2008 to 2022","result":"no elephant deaths on the railway tracks in the last 22 months (as of 21 December 2025)","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.newindianexpress.com/states/tamil-nadu/2025/Dec/21/no-elephant-deaths-on-railway-tracks-in-madukkarai-stretch-in-22-months","https://indianmasterminds.com/feature-stories-on-bureaucrats-changemakers/ai-saving-elephants-tamil-nadu-railway-tracks-supriya-sahu-218109/","https://www.newindianexpress.com/states/tamil-nadu/2026/Jul/08/ai-camera-system-along-railway-tracks-ends-jumbo-deaths-in-madukkarai-forest-range","https://www.indiatoday.in/trending-news/story/tamil-nadu-forest-department-shares-video-showing-how-ai-ensures-safety-of-elephants-2770278-2025-08-12","https://www.dtnext.in/news/tamilnadu/ai-powered-elephant-alert-system-enables-9481-safe-crossings-records-zero-train-track-deaths","https://www.thehindu.com/news/cities/Coimbatore/trial-run-of-ai-based-early-warning-system-near-coimbatore-proves-promising-as-field-staff-save-elephants/article67716099.ece","https://www.downtoearth.org.in/wildlife-biodiversity/ai-enabled-system-successful-in-preventing-elephant-train-collisions-on-kerala-tamil-nadu-border","https://www.newindianexpress.com/states/kerala/2026/Aug/13/elephant-calf-killed-by-train-near-keralas-walayar-traffic-disrupted-for-hours","https://timesofindia.indiatimes.com/city/thiruvananthapuram/railways-yet-to-find-solution-to-jumbo-deaths-on-rail-track/articleshow/133249199.cms","https://www.thehindu.com/news/national/kerala/anec-serves-legal-notice-over-elephant-deaths-on-walayar-rail-corridor/article71410593.ece"],"source_refs":["AIBD-0101-source-0002","AIBD-0101-source-0003","AIBD-0101-source-0004","AIBD-0101-source-0005","AIBD-0101-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the same 7.05 km Madukkarai railway stretch before the cameras (2008 to 2022)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: Eleven wild elephants were killed on railway tracks in Madukkarai stretch from 2008 to 2022. Result: no elephant deaths on the railway tracks in the last 22 months (as of 21 December 2025).","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"about 9,481 safe elephant crossings and no elephant deaths on the stretch","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0101-reported-measure","evaluation_data":null},{"case_id":"AIBD-0567","case_url":"records/AIBD-0567.html","measure":"Average rating of the AI-generated content out of 10.","reported_result":"Average rating of 8 out of 10.","baseline":"Manual drafting of each advert, with varied formatting","result":"A substantial reduction in drafting time, reported by managers without a figure, and uniform formatting","baseline_value":null,"result_value":null,"unit":null,"effect":"Average rating of 8 out of 10.","design":"testimonial.","sample":"Senior and recruiting managers at West Berkshire Council; sample size not stated.","uncertainty":"Sample size, survey date and question wording are not reported. No measured drafting time is given.","comparison_basis":"assumed","source_urls":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/west-berkshire-council-development-generative-ai-tools","https://www.local.gov.uk/case-studies/west-berkshire-council-development-generative-ai-tools","https://web.archive.org/web/2026/https://www.ukauthority.com/articles/a-starting-point-for-ai-in-citizen-services/"],"source_refs":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/west-berkshire-council-development-generative-ai-tools"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"None stated.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"testimonial.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Average rating of 8 out of 10.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"testimonial.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Survey of senior managers who reviewed AI-generated job adverts.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Average rating of the AI-generated content out of 10.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Senior and recruiting managers at West Berkshire Council; sample size not stated.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sample size, survey date and question wording are not reported. No measured drafting time is given.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0567-reported-measure","evaluation_data":null},{"case_id":"AIBD-0557","case_url":"records/AIBD-0557.html","measure":"90% less time and cost","reported_result":"90% less time and cost","baseline":"Time and cost of selected data-retrieval tasks in the existing system (figures not published)","result":"90 percent less time and cost for those tasks in the pilot","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/north-yorkshire-council-using-ai-reimagine-childrens-social-care","https://www.local.gov.uk/case-studies/north-yorkshire-council-using-ai-reimagine-childrens-social-care","https://www.globalgovernmentforum.com/slick-cities-how-local-authorities-are-using-ai-to-tackle-their-most-pressing-problems/"],"source_refs":["AIBD-0557-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0557-reported-measure","evaluation_data":null},{"case_id":"AIBD-0549","case_url":"records/AIBD-0549.html","measure":null,"reported_result":null,"baseline":"66 per cent","result":"100 per cent","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/hertfordshire-county-council-ai-road-maintenance-robot","https://www.local.gov.uk/case-studies/hertfordshire-county-council-ai-road-maintenance-robot","https://www.bbc.co.uk/news/articles/cy8x8gxkmk3o","https://www.rac.co.uk/drive/news/motoring-news/worlds-first-pothole-preventing-robot-using-ai-to-tackle-growing-problem/"],"source_refs":["AIBD-0549-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0549-reported-measure","evaluation_data":null},{"case_id":"AIBD-0169","case_url":"records/AIBD-0169.html","measure":"reduced leakage run rate times by up to 40%","reported_result":"reduced leakage run rate times by up to 40%","baseline":"a daily manual check of pressure and flow deviations, with leak run times not quantified","result":"run time of critical burst mains cut by 40%","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.haskoning.com/en/projects/ses-water-achieves-a-reduction-in-leak-run-time","https://www.bbc.co.uk/news/articles/c4n5pwny536o"],"source_refs":["AIBD-0169-source-0001","AIBD-0169-source-0002","AIBD-0169-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"leak run times before the sensor network","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: previous leakage run rate times (not quantified). Result: reduce our leakage run rate times by up to 40%.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"reduced leakage run rate times by up to 40%","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0169-reported-measure","evaluation_data":null},{"case_id":"AIBD-0319","case_url":"records/AIBD-0319.html","measure":"transcribes te reo Māori with 92% accuracy","reported_result":"transcribes te reo Māori with 92% accuracy","baseline":null,"result":"an accuracy of 92%","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.rnz.co.nz/media/226","https://www.auckland.ac.nz/en/news/2024/03/18/te-reo-artificial-intelligence.html","https://www.technologyreview.com/2022/04/22/1050394/artificial-intelligence-for-the-people/","https://blogs.nvidia.com/blog/te-hiku-media-maori-speech-ai/"],"source_refs":["AIBD-0319-source-0001","AIBD-0319-source-0002","AIBD-0319-source-0003","AIBD-0319-source-0004","AIBD-0319-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: te reo Māori speech recognition attempts by major tech companies","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Result: an accuracy of 92%.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"transcribes te reo Māori with 92% accuracy","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0319-reported-measure","evaluation_data":null},{"case_id":"AIBD-0306","case_url":"records/AIBD-0306.html","measure":"2.76 out of 5 for satisfaction and 2.43 out of 4 for trust on average","reported_result":"2.76 out of 5 for satisfaction and 2.43 out of 4 for trust on average","baseline":null,"result":"She was happy to get an answer without having to call her husband.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://arxiv.org/html/2403.15604","https://arxiv.org/abs/2503.05899"],"source_refs":["AIBD-0306-source-0001","AIBD-0306-source-0002","AIBD-0306-source-0003","AIBD-0306-source-0004","AIBD-0306-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"asking a sighted family member (P1's husband)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Result: She was happy to get an answer without having to call her husband.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"2.76 out of 5 for satisfaction and 2.43 out of 4 for trust on average","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0306-reported-measure","evaluation_data":null},{"case_id":"AIBD-0464","case_url":"records/AIBD-0464.html","measure":null,"reported_result":null,"baseline":"FAHR describes manual HR processes as time-consuming, inconsistent and not personalized.","result":"No measured change is reported.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-489","https://apps.apple.com/ae/app/fahr/id918247230","https://apps.apple.com/us/app/fahr/id918247230","https://www.fahr.gov.ae/en/news/fahr-wins-18-stevie-awards-in-2026/"],"source_refs":["AIBD-0464-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0464-reported-measure","evaluation_data":null},{"case_id":"AIBD-0139","case_url":"records/AIBD-0139.html","measure":"drove Max Paul Franklin 13 miles to a VA emergency room","reported_result":"drove Max Paul Franklin 13 miles to a VA emergency room","baseline":null,"result":"reached the VA emergency room 13 miles away, on his account without intervening","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.hindustantimes.com/world-news/us-news/elon-musk-responds-to-tesla-full-self-driving-that-drove-a-man-to-the-hospital-saving-his-life-101712894057291.html","https://x.com/MAXPAULFRANKLIN/status/1777918962870718909","https://thedriven.io/2024/04/12/autonomous-ev-saves-man-tesla-model-y-drives-heart-attack-victim-20-kms-to-hospital/","https://www.ilfattoquotidiano.it/2024/04/14/ha-un-infarto-e-crisi-glicemica-mentre-e-al-volante-la-sua-tesla-gli-salva-la-vita-ha-guidato-da-sola-per-20-km-fino-al-pronto-soccorso/7513224/","https://www.teslaoracle.com/2024/04/12/the-true-story-of-how-fsd-v12-helped-a-tesla-owner-reach-hospital-and-saved-his-life/","https://www.siasat.com/teslas-full-self-driving-feature-saves-mans-life-in-us-musk-responds-3007838","https://static.nhtsa.gov/odi/inv/2024/INIM-PE24031-27346.pdf"],"source_refs":["AIBD-0139-source-0001","AIBD-0139-source-0002","AIBD-0139-source-0003","AIBD-0139-source-0004","AIBD-0139-source-0005","AIBD-0139-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"drove Max Paul Franklin 13 miles to a VA emergency room","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"drove Max Paul Franklin 13 miles to a VA emergency room","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0139-reported-measure","evaluation_data":null},{"case_id":"AIBD-0626","case_url":"records/AIBD-0626.html","measure":"AUROC and sepsis rate by risk category.","reported_result":"an area under the ROC curve of 0.81","baseline":"Goal of 0.75","result":"0.81 (forced majority adjudication)","baseline_value":null,"result_value":null,"unit":null,"effect":"AUROC 0.81; sepsis 3.02% low to 69.70% very high.","design":"other.","sample":"746 patients with suspected sepsis at three US hospitals.","uncertainty":"Retrospective; same-site adjudicators; mechanical ventilation endpoint not met.","comparison_basis":"measured","source_urls":["https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN230036.pdf"],"source_refs":["AIBD-0626-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Physician adjudication of Sepsis-3.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"AUROC 0.81; sepsis 3.02% low to 69.70% very high.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One score per patient.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective study of prospectively collected data and samples; sepsis adjudicated by three physicians blinded to the score.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"AUROC and sepsis rate by risk category.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"None public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"746 patients with suspected sepsis at three US hospitals.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective; same-site adjudicators; mechanical ventilation endpoint not met.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0626-reported-measure","evaluation_data":null},{"case_id":"AIBD-0370","case_url":"records/AIBD-0370.html","measure":"Adenoma detection rate.","reported_result":"a 1 to 1.5 cm adenoma that biopsy confirmed as stage I colorectal cancer","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":"61.5% with AI against 69.8% without AI (p = 0.097); advanced neoplasia 37.9% against 36.3%.","design":"randomized.","sample":"361 people aged 50 to 74 with a positive fecal immunochemical test, one hospital in Palencia, Spain, October 2023 to February 2025.","uncertainty":"Single center with high baseline detection rates; the authors call for multicenter studies.","comparison_basis":null,"source_urls":["https://health.ltn.com.tw/article/breakingnews/4646402","https://tw.news.yahoo.com/%E8%81%96%E9%A6%AC%E7%88%BE%E5%AE%9A%E9%86%AB%E9%99%A2%E9%81%8B%E7%94%A8ai%E5%85%A7%E8%A6%96%E9%8F%A1%E6%AA%A2%E6%9F%A5-%E5%A4%A7%E8%85%B8%E7%99%8C%E7%84%A1%E6%89%80%E9%81%81%E5%BD%A2-052431696.html","https://tw.news.yahoo.com/ai%E5%85%A7%E8%A6%96%E9%8F%A1%E6%AA%A2%E6%9F%A5-%E5%A4%A7%E8%85%B8%E7%99%8C%E7%84%A1%E6%89%80%E9%81%81%E5%BD%A2-113000020.html","https://oecd.ai/en/incidents/2024-04-19-de89","https://www.stm.org.tw/health/list_detail.aspx?TypeID=87&ids=958","https://pubmed.ncbi.nlm.nih.gov/41854124/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11038826/"],"source_refs":["AIBD-0370-source-0006"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"No significant gain in detection; CADx specificity was 66.7% against 51.9% for optical diagnosis by endoscopists.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Conventional colonoscopy without AI.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"61.5% with AI against 69.8% without AI (p = 0.097); advanced neoplasia 37.9% against 36.3%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"CAD EYE-assisted colonoscopy.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Adenoma detection rate and advanced neoplasia detection rate by colonoscopy and pathology.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Adenoma detection rate.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"361 people aged 50 to 74 with a positive fecal immunochemical test, one hospital in Palencia, Spain, October 2023 to February 2025.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Single center with high baseline detection rates; the authors call for multicenter studies.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0370-reported-measure","evaluation_data":null},{"case_id":"AIBD-0216","case_url":"records/AIBD-0216.html","measure":"Share of sessions in which the student passed the exit ticket.","reported_result":"4 percentage points more likely to master topics","baseline":"62% student passing rate on exit tickets (control tutors)","result":"66% passing rate (tutors with Tutor CoPilot), 4 percentage points higher","baseline_value":null,"result_value":null,"unit":null,"effect":"4 percentage points higher (62% to 66%); 9 points for students of lower-rated tutors.","design":"randomized.","sample":"900 tutors at FEV Tutor randomized 450 to each arm, of whom 783 tutored students in the study; 1,013 students in grades 3 to 6 from nine schools of a large southern U.S. district; 27 March to 17 May 2024.","uncertainty":"No significant gain on end-of-year test scores; seven weeks of exposure; tutors were randomized while students were not; the authors built the tool and wrote the study; the work remains a preprint.","comparison_basis":"measured","source_urls":["https://arxiv.org/pdf/2410.03017v2","https://github.com/rosewang2008/tutor-copilot/blob/25c6eab7e416614f2033358bceddd30a4d5ca616/Tutor_CoPilot_Demo.ipynb","https://www.edweek.org/technology/what-happens-when-an-ai-assistant-helps-the-tutor-instead-of-the-student/2024/10","https://www.the74million.org/article/study-ai-assisted-tutoring-boosts-students-math-skills/","https://www.technologyreview.com/2024/10/28/1106251/this-ai-system-makes-human-tutors-better-at-teaching-children-math/","https://www.the74million.org/article/tutoring-giants-sudden-demise-linked-to-end-of-federal-relief-funds/","https://arxiv.org/pdf/2410.03017v3"],"source_refs":["AIBD-0216-source-0007"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Intent-to-treat effect of 4 percentage points; 14 points estimated for sessions in which the tool was used.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Students of control tutors without access to the tool.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"4 percentage points higher (62% to 66%); 9 points for students of lower-rated tutors.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"About seven weeks of access during live chat sessions.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Session exit-ticket results from the tutoring platform; end-of-year NWEA MAP math scores were the initial preregistered outcome; classifiers applied to more than 350,000 chat messages.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Share of sessions in which the student passed the exit ticket.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Preregistration on the Open Science Framework and a demonstration notebook on GitHub.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"900 tutors at FEV Tutor randomized 450 to each arm, of whom 783 tutored students in the study; 1,013 students in grades 3 to 6 from nine schools of a large southern U.S. district; 27 March to 17 May 2024.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"No significant gain on end-of-year test scores; seven weeks of exposure; tutors were randomized while students were not; the authors built the tool and wrote the study; the work remains a preprint.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0216-reported-measure","evaluation_data":null},{"case_id":"AIBD-0232","case_url":"records/AIBD-0232.html","measure":"jailed Hughes (18) for 28 months and Potter (19) for 34 months","reported_result":"jailed Hughes (18) for 28 months and Potter (19) for 34 months","baseline":"Hughes was not known in the Northwich area","result":"we were quickly able to identify him as a suspect","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.deeside.com/facial-recognition-technology-helps-to-identify-deeside-drug-dealer","https://www.cheshire.police.uk/news/cheshire/news/articles/2024/6/facial-recognition-technology-helps-to-identify-northwich-drug-dealers/","https://www.leaderlive.co.uk/news/24424400.county-lines-drug-dealers-operating-northwich-jailed/","https://oecd.ai/en/incidents/2024-07-02-ee42"],"source_refs":["AIBD-0232-source-0001","AIBD-0232-source-0002","AIBD-0232-source-0003","AIBD-0232-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: conventional identification of a suspect who was not known locally","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: Hughes was not known in the Northwich area. Result: we were quickly able to identify him as a suspect.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"jailed Hughes (18) for 28 months and Potter (19) for 34 months","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0232-reported-measure","evaluation_data":null},{"case_id":"AIBD-0213","case_url":"records/AIBD-0213.html","measure":"standardized assessment score (SD).","reported_result":"0.31 standard deviations overall and 0.23 in English","baseline":"671 students randomized to regular instruction","result":"657 students offered twelve 90-minute Copilot sessions; overall assessment 0.310 SD higher","baseline_value":null,"result_value":null,"unit":null,"effect":"0.310 SD overall (95% CI 0.176 to 0.444; n = 654); 0.238 SD in English; 0.309 SD in AI knowledge; 0.206 SD in school examinations.","design":"randomized.","sample":"Nine public senior secondary schools in Benin City, Nigeria; 1,328 randomized volunteers, 657 offered the program and 671 controls.","uncertainty":"Attrition at endline (422 treated and 337 controls finished the final assessment in the original paper); short-term outcomes only; the digital-skills effect is imprecise (p = 0.066).","comparison_basis":"measured","source_urls":["https://doi.org/10.3390/educsci16111780","https://openknowledge.worldbank.org/bitstreams/8fcccf10-83c9-4295-b01c-5a68aac83b06/download","https://hdl.handle.net/10986/43212","https://eric.ed.gov/?id=ED676624","https://blogs.worldbank.org/en/education/From-chalkboards-to-chatbots-Transforming-learning-in-Nigeria","https://blogs.worldbank.org/en/education/From-chalkboards-to-chatbots-in-Nigeria"],"source_refs":["AIBD-0213-source-0006"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated: the estimate covers Copilot together with curriculum-aligned activities, pair work, teacher guidance and extra lab time.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"regular instruction.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"0.310 SD overall (95% CI 0.176 to 0.444; n = 654); 0.238 SD in English; 0.309 SD in AI knowledge; 0.206 SD in school examinations.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Up to twelve 90-minute after-school sessions over six weeks, June and July 2024.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Endline assessment covering English, knowledge of AI and digital skills, plus school examination scores.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"standardized assessment score (SD).","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"World Bank reproducibility package by the study authors, doi 10.60572/cre7-j348.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Nine public senior secondary schools in Benin City, Nigeria; 1,328 randomized volunteers, 657 offered the program and 671 controls.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Attrition at endline (422 treated and 337 controls finished the final assessment in the original paper); short-term outcomes only; the digital-skills effect is imprecise (p = 0.066).","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0213-reported-measure","evaluation_data":null},{"case_id":"AIBD-0385","case_url":"records/AIBD-0385.html","measure":"over 13 million users","reported_result":"over 13 million users","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0385-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0385-reported-measure","evaluation_data":null},{"case_id":"AIBD-0657","case_url":"records/AIBD-0657.html","measure":"Upper-air and 2-meter temperature skill; event discrimination; cyclone track and intensity error.","reported_result":"Pangu-Weather discriminated events better than the IFS","baseline":"IFS discrimination of summer 2-meter temperature events over Europe","result":"Significantly better discrimination by Pangu-Weather for thresholds between the 75th and 90th percentiles","baseline_value":null,"result_value":null,"unit":null,"effect":"Comparable global skill; better summer event discrimination for the 75th to 90th percentiles; similar cyclone track error; worse cyclone intensity.","design":"controlled observational.","sample":"Ten-day forecasts from 1 March 2022 to 28 February 2023, with 2018 for tropical cyclones; Northern Hemisphere and Europe.","uncertainty":"Deterministic forecasts from one model; smooth fields and growing bias; poor cyclone intensity; no measure of use by forecasters.","comparison_basis":"measured","source_urls":["https://journals.ametsoc.org/view/journals/bams/105/6/BAMS-D-23-0162.1.xml"],"source_refs":["AIBD-0657-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Comparable accuracy at much lower computing cost; cost not quantified in the paper.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Operational IFS forecasts from the same initial conditions.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Comparable global skill; better summer event discrimination for the 75th to 90th percentiles; similar cyclone track error; worse cyclone intensity.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Research runs only.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Standard verification scores (correlation, RMSE, area under the ROC curve, cyclone position and pressure error) against IFS analyses, synoptic observations and IBTrACS.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Upper-air and 2-meter temperature skill; event discrimination; cyclone track and intensity error.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Model weights public for noncommercial use; ECMWF toolkit public; ECMWF forecasts available.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Ten-day forecasts from 1 March 2022 to 28 February 2023, with 2018 for tropical cyclones; Northern Hemisphere and Europe.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Deterministic forecasts from one model; smooth fields and growing bias; poor cyclone intensity; no measure of use by forecasters.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0657-reported-measure","evaluation_data":null},{"case_id":"AIBD-0006","case_url":"records/AIBD-0006.html","measure":"Mean change in forced vital capacity in mL at 12 weeks.","reported_result":"mean forced vital capacity change of +98.4 mL versus -20.3 mL on placebo over 12 weeks","baseline":"−20.3 ml (95% confidence interval −116.1 to 75.6) for the placebo group","result":"+98.4 ml (95% confidence interval 10.9 to 185.9) for patients in the 60 mg rentosertib QD group","baseline_value":null,"result_value":null,"unit":null,"effect":"+98.4 mL (95% CI 10.9 to 185.9) on 60 mg once daily versus -20.3 mL (95% CI -116.1 to 75.6) on placebo.","design":"randomized.","sample":"71 adults with idiopathic pulmonary fibrosis at 21 sites in China, randomized 1:1:1:1 to three doses or placebo.","uncertainty":"Small groups of about 18 patients; 12 weeks only; 12 of 18 patients in the 60 mg group completed the placebo-controlled period. Seven patients on rentosertib stopped treatment because of liver toxicity, and three patients on 60 mg had an acute exacerbation of IPF against one on placebo. The sponsor ran the trial and eleven authors are its employees.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12353801/","https://clinicaltrials.gov/study/NCT05938920","https://pmc.ncbi.nlm.nih.gov/articles/PMC13330183/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13304925/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13634690/","https://www.healio.com/news/pulmonology/20250707/qa-aiestablished-drug-breakthrough-in-the-treatment-of-ipf","https://insilico.com/news/xmjsn4l091-insilico-initiates-phase-iii-clinical-tr","https://www.news-medical.net/news/20260910/Generative-AI-driven-drug-Rentosertib-enters-Phase-III-trial-for-idiopathic-pulmonary-fibrosis.aspx"],"source_refs":["AIBD-0006-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not measured. The trial tests the drug; it does not compare AI-led discovery with any other way of finding a drug.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Placebo once daily, with standard care continued in all groups.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"+98.4 mL (95% CI 10.9 to 185.9) on 60 mg once daily versus -20.3 mL (95% CI -116.1 to 75.6) on placebo.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Oral rentosertib for 12 weeks.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Spirometry; change in forced vital capacity from baseline to week 12, a secondary efficacy measure. Safety was the primary endpoint.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean change in forced vital capacity in mL at 12 weeks.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"71 adults with idiopathic pulmonary fibrosis at 21 sites in China, randomized 1:1:1:1 to three doses or placebo.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small groups of about 18 patients; 12 weeks only; 12 of 18 patients in the 60 mg group completed the placebo-controlled period. Seven patients on rentosertib stopped treatment because of liver toxicity, and three patients on 60 mg had an acute exacerbation of IPF against one on placebo. The sponsor ran the trial and eleven authors are its employees.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0006-reported-measure","evaluation_data":null},{"case_id":"AIBD-0161","case_url":"records/AIBD-0161.html","measure":"helicopters were dispatched within 30 minutes and the fire was held to three acres","reported_result":"helicopters were dispatched within 30 minutes and the fire was held to three acres","baseline":"anyone else calling 911","result":"within 10 minutes of detecting smoke, 10 to 12 minutes","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.denver7.com/news/wildfire/douglas-county-wildfire-first-detected-by-cameras-equipped-with-ai","https://core.coop/core-electric-cooperative-and-xcel-energy-ai-powered-cameras-aid-in-bear-creek-fire-containment/","https://www.renewableenergyworld.com/power-grid/outage-management/ai-cameras-prevented-large-wildfire-utilities-say/","https://oecd.ai/en/incidents/2023-11-07-64f3","https://www.cbsnews.com/colorado/news/ai-powered-fire-spotting-cameras-dispatch-firefighters-wildfire/","https://www.cbsnews.com/colorado/news/colorado-county-uses-fire-suppression-heliteam-ai-to-combat-wildfires/"],"source_refs":["AIBD-0161-source-0002","AIBD-0161-source-0003","AIBD-0161-source-0004","AIBD-0161-source-0005","AIBD-0161-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: members of the public calling 911","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: anyone else calling 911. Result: within 10 minutes of detecting smoke, 10 to 12 minutes.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"helicopters were dispatched within 30 minutes and the fire was held to three acres","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0161-reported-measure","evaluation_data":null},{"case_id":"AIBD-0242","case_url":"records/AIBD-0242.html","measure":"Odds ratio of a crash involving an automated driving system against a human-driven vehicle under a given condition.","reported_result":"In rain, a Level 4 automated vehicle had 0.335 times the crash odds of a human-driven vehicle","baseline":"Human-Driven Vehicles (reference odds)","result":"the odds of an ADS accident occurring in rainy weather are 0.335 times; 5.25 times higher at dawn/dusk and 1.98 times higher when turning","baseline_value":null,"result_value":null,"unit":null,"effect":"Lower odds for automated driving systems in most matched scenarios, such as 0.335 in rain; higher odds at dawn or dusk (5.250) and when turning (1.988).","design":"controlled observational.","sample":"2,100 automated-vehicle crash records (California DMV and NHTSA Standing General Order; 1,099 SAE Level 4 and 1,001 Level 2) and 35,113 human-driven crashes from California's SWITRS, with automated-vehicle records up to June 2023. The matched model uses only California Level 4 crashes: 548 strata at a sampling ratio of 1 to 5.","uncertainty":"The data hold crashes only, so the odds compare crash conditions and do not give crash rates. The authors write that direct comparison of crash numbers is not viable because exposure and fleet size differ so much. They also could not analyze which automation level or system was active in a crash.","comparison_basis":"measured","source_urls":["https://www.ebi.ac.uk/europepmc/webservices/rest/PMC11189485/fullTextXML","https://europepmc.org/article/MED/38890354","https://pmc.ncbi.nlm.nih.gov/articles/PMC11189485/"],"source_refs":["AIBD-0242-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Human-driven vehicle crashes in matched scenarios.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Lower odds for automated driving systems in most matched scenarios, such as 0.335 in rain; higher odds at dawn or dusk (5.250) and when turning (1.988).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"No vehicle miles traveled for automated vehicles; the authors add annual average daily traffic by road type for a descriptive comparison.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Matched case-control conditional logistic regression on police and regulator crash reports. Automated-vehicle crashes are the cases and human-driven crashes the controls, matched at the same location and road type, with road, environment, pre-crash movement, crash type and outcome as explanatory variables.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Odds ratio of a crash involving an automated driving system against a human-driven vehicle under a given condition.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Public SWITRS human-driven crash data; supplementary information and source data with the article; code available from the authors on request.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"2,100 automated-vehicle crash records (California DMV and NHTSA Standing General Order; 1,099 SAE Level 4 and 1,001 Level 2) and 35,113 human-driven crashes from California's SWITRS, with automated-vehicle records up to June 2023. The matched model uses only California Level 4 crashes: 548 strata at a sampling ratio of 1 to 5.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The data hold crashes only, so the odds compare crash conditions and do not give crash rates. The authors write that direct comparison of crash numbers is not viable because exposure and fleet size differ so much. They also could not analyze which automation level or system was active in a crash.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0242-reported-measure","evaluation_data":null},{"case_id":"AIBD-0658","case_url":"records/AIBD-0658.html","measure":"Hit rate (actives per molecules tested) and affinity.","reported_result":"a hit rate of 54%, against 51% for the crystal structure","baseline":"23% hit rate (51 of 223 molecules)","result":"26% hit rate (42 of 161 molecules)","baseline_value":null,"result_value":null,"unit":null,"effect":"54% against 51% (sigma-2); 26% against 23% (5-HT2A); not significantly different.","design":"controlled observational.","sample":"About 500 synthesized molecules from docking 490 million (sigma-2) and 1.6 billion (5-HT2A) molecules.","uncertainty":"Two receptors; functional hit rates for 5-HT2A were 1 to 5%; retrospective enrichment favored experimental structures.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11253030/"],"source_refs":["AIBD-0658-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Predicted structures matched experimental ones for ligand discovery.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Docking against experimental crystal and cryo-EM structures.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"54% against 51% (sigma-2); 26% against 23% (5-HT2A); not significantly different.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Docking campaigns at UCSF; assays at UNC and elsewhere.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Radioligand binding assays of docked molecules; hit rate and affinity.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Hit rate (actives per molecules tested) and affinity.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Models, docking results and compounds described in the supplement; ZINC library public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"About 500 synthesized molecules from docking 490 million (sigma-2) and 1.6 billion (5-HT2A) molecules.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two receptors; functional hit rates for 5-HT2A were 1 to 5%; retrospective enrichment favored experimental structures.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0658-reported-measure","evaluation_data":null},{"case_id":"AIBD-0663","case_url":"records/AIBD-0663.html","measure":"Correct classifications, false negatives, false positives; cost over three years.","reported_result":"eVorta correctly classified 482 of 492 cat images (97.9%) against 469 (95.3%) for staff","baseline":"Staff missed 11 of 68 cat sequences","result":"eVorta missed one of 68 cat sequences","baseline_value":null,"result_value":null,"unit":null,"effect":"97.9% against 95.3% of cat images; one against 11 missed sequences; A$2.15 million estimated saving with 4G.","design":"controlled observational.","sample":"101,586 images from 42 cameras, 11 November 2020 to 9 September 2021; 492 cat and 155 echidna images.","uncertainty":"2,751 false cat alerts against five by staff; costs modeled; one program; written by the deployer, with no eVorta author or funding.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11196731/"],"source_refs":["AIBD-0663-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Fewer missed cats; 81% lower estimated monitoring cost with 4G cameras.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Experienced field staff.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"97.9% against 95.3% of cat images; one against 11 missed sequences; A$2.15 million estimated saving with 4G.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Program monitoring cameras.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Image and sequence classification compared between eVorta and staff; cost and emissions from a decision tool with program inputs.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Correct classifications, false negatives, false positives; cost over three years.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Decision tool in the supplement.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"101,586 images from 42 cameras, 11 November 2020 to 9 September 2021; 492 cat and 155 echidna images.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"2,751 false cat alerts against five by staff; costs modeled; one program; written by the deployer, with no eVorta author or funding.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0663-reported-measure","evaluation_data":null},{"case_id":"AIBD-0315","case_url":"records/AIBD-0315.html","measure":"achieving around 70% accuracy already","reported_result":"achieving around 70% accuracy already","baseline":"days-long manual assessments","result":"hours","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.halotrust.org/news/prince-harry-trials-cutting-edge-tech-in-ukraine/","http://web.archive.org/web/20260520220522/https://www.halotrust.org/news/artificial-intelligence-to-help-accelerate-landmine-clearance/","https://www.halotrust.org/news/using-drones-and-ai-to-eradicate-landmines/","https://kelo.com/2026/04/30/man-machine-and-ai-combine-in-battle-to-clear-ukraine-of-mines/"],"source_refs":["AIBD-0315-source-0001","AIBD-0315-source-0002","AIBD-0315-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: manual analyst review of drone imagery of a minefield","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: three to five days to look through the imagery for an average minefield.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"achieving around 70% accuracy already","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0315-reported-measure","evaluation_data":null},{"case_id":"AIBD-0070","case_url":"records/AIBD-0070.html","measure":"IMO points out of 42.","reported_result":"awarded 28 of 42 points, a silver-medal score","baseline":"gold-medal threshold of 29 points, reached by 58 of 609 contestants","result":"28 of 42 points, one point below the gold threshold and within the silver range","baseline_value":null,"result_value":null,"unit":null,"effect":"28 of 42 points from four fully solved problems, one point below the gold threshold of 29.","design":"case study.","sample":"The six IMO 2024 problems; five formalized in Lean by experts and one geometry problem handled by AlphaGeometry 2.","uncertainty":"Solutions took up to three days against the contest's two sessions of 4.5 hours, people translated the problems into Lean, the IMO did not certify the score, and the graders were engaged by the developer.","comparison_basis":"measured","source_urls":["https://deepmind.google/discover/blog/ai-solves-imo-problems-at-silver-medal-level/","https://www.scientificamerican.com/article/ai-reaches-silver-medal-level-at-this-years-math-olympiad/","https://www.technologyreview.com/2024/07/25/1095315/google-deepminds-ai-systems-can-now-solve-complex-math-problems/","https://www.nature.com/articles/d41586-024-02441-2","https://www.nature.com/articles/d41586-025-02343-x","https://deepmind.google/discover/blog/advanced-version-of-gemini-with-deep-think-officially-achieves-gold-medal-standard-at-the-international-mathematical-olympiad/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12999475/"],"source_refs":["AIBD-0070-source-0007"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Official IMO 2024 medal thresholds for human contestants.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"28 of 42 points from four fully solved problems, one point below the gold threshold of 29.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Up to three days of computation per problem.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"The Lean kernel checked the AlphaProof proofs, and two former IMO gold medalists engaged by DeepMind scored all solutions under IMO point rules.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"IMO points out of 42.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"DeepMind's 2024 blog links to the system's IMO 2024 solutions, and the Nature paper shows the AlphaProof proofs in its Extended Data figures.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"The six IMO 2024 problems; five formalized in Lean by experts and one geometry problem handled by AlphaGeometry 2.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Solutions took up to three days against the contest's two sessions of 4.5 hours, people translated the problems into Lean, the IMO did not certify the score, and the graders were engaged by the developer.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0070-reported-measure","evaluation_data":null},{"case_id":"AIBD-0548","case_url":"records/AIBD-0548.html","measure":null,"reported_result":null,"baseline":"Information and support available during traditional service hours","result":"Information available at any hour through the app and website","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["http://web.archive.org/web/20251218173213/https://www.local.gov.uk/case-studies/halton-borough-council-digital-front-door-family-hub-services-0","https://www.local.gov.uk/case-studies/halton-borough-council-digital-front-door-family-hub-services-0"],"source_refs":["AIBD-0548-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0548-reported-measure","evaluation_data":null},{"case_id":"AIBD-0609","case_url":"records/AIBD-0609.html","measure":"Mean EHR time and mean time in notes per appointment.","reported_result":"15.9% less time in notes and 8.5% less time in the EHR","baseline":"Nonuser mean time in notes","result":"15.9% lower mean time in notes (about 1.8 minutes per appointment)","baseline_value":null,"result_value":null,"unit":null,"effect":"-8.5% EHR time (95% CI -12.8% to -3.9%); -15.9% note time (95% CI -21.2% to -10.4%).","design":"controlled observational.","sample":"125 scribe users and 478 nonusers, University of Chicago ambulatory clinics.","uncertainty":"Voluntary users, one site, three months, no change in after-hours time.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12514625/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13154711/"],"source_refs":["AIBD-0609-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"About 1.8 minutes less in notes per appointment.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Propensity-weighted nonusers.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"-8.5% EHR time (95% CI -12.8% to -3.9%); -15.9% note time (95% CI -21.2% to -10.4%).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Pilot users across specialties.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"EHR audit-log time per appointment.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean EHR time and mean time in notes per appointment.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Data sharing statement in Supplement 2.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"125 scribe users and 478 nonusers, University of Chicago ambulatory clinics.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Voluntary users, one site, three months, no change in after-hours time.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0609-reported-measure","evaluation_data":null},{"case_id":"AIBD-0619","case_url":"records/AIBD-0619.html","measure":"Post-test score.","reported_result":"the ChatGPT-4 group scored 22.8 against 19.24 (p = 0.005, Cohen's d = 0.58)","baseline":"Mean post-test score 19.24 (control)","result":"Mean post-test score 22.8 (ChatGPT-4)","baseline_value":null,"result_value":null,"unit":null,"effect":"22.8 against 19.24 (Cohen's d = 0.58, p = 0.005).","design":"randomized.","sample":"100 fourth-year medical students, Islamic International Medical College, February to July 2024.","uncertainty":"Single institution; one topic; short-term outcome; two-hour exposure; no formal washout between crossover phases; scores were similar after crossover.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC13525506/"],"source_refs":["AIBD-0619-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The difference is attributed to the ChatGPT-4 learning pathways.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Conventional educational resources.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"22.8 against 19.24 (Cohen's d = 0.58, p = 0.005).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two hours of study on one topic, then one hour with the other method after crossover.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Topic-specific pre-test and post-test.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Post-test score.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"100 fourth-year medical students, Islamic International Medical College, February to July 2024.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Single institution; one topic; short-term outcome; two-hour exposure; no formal washout between crossover phases; scores were similar after crossover.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0619-reported-measure","evaluation_data":null},{"case_id":"AIBD-0669","case_url":"records/AIBD-0669.html","measure":"Minutes per week on lesson and resource preparation.","reported_result":"56.2 minutes a week on Year 7 and 8 science lesson preparation against 81.5 minutes","baseline":"81.5 minutes per week","result":"56.2 minutes per week","baseline_value":null,"result_value":null,"unit":null,"effect":"56.2 against 81.5 minutes per week (69%).","design":"randomized.","sample":"259 Year 7 and 8 science teachers in 68 schools.","uncertainty":"Self-reported diary times; quality reviewed for 30 teachers; one subject and two year groups; free ChatGPT version changed during the trial.","comparison_basis":"measured","source_urls":["https://www.nfer.ac.uk/publications/chatgpt-in-lesson-preparation-a-teacher-choices-trial/","https://d2tic4wvo1iusb.cloudfront.net/production/documents/projects/chatgpt_in_lesson_planning_-_evaluation_report.pdf","https://web.archive.org/web/2025/https://schoolsweek.co.uk/using-chatgpt-can-cut-lesson-planning-by-31-eef-study/"],"source_refs":["AIBD-0669-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"25.3 minutes saved per week.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Teachers asked not to use generative AI.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"56.2 against 81.5 minutes per week (69%).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Ten weeks of ChatGPT use with a guide.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Teacher-reported weekly preparation time; blind expert panel review of sampled resources.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Minutes per week on lesson and resource preparation.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Full report published by the EEF.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"259 Year 7 and 8 science teachers in 68 schools.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Self-reported diary times; quality reviewed for 30 teachers; one subject and two year groups; free ChatGPT version changed during the trial.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0669-reported-measure","evaluation_data":null},{"case_id":"AIBD-0671","case_url":"records/AIBD-0671.html","measure":"Standardized solution quality.","reported_result":"raised solution quality by 0.37 SD and spent 16.4% less time","baseline":"Control individuals; teams without AI scored 0.24 SD above them","result":"Individuals with AI 0.37 SD above; teams with AI 0.39 SD above","baseline_value":null,"result_value":null,"unit":null,"effect":"0.37 SD (individuals with AI), 0.39 SD (teams with AI), 0.24 SD (teams without AI).","design":"randomized.","sample":"776 P&G professionals in R&D and commercial roles.","uncertainty":"One-day workshop tasks; quality judged as proposals; P&G funding and co-authors.","comparison_basis":"measured","source_urls":["https://www.nber.org/system/files/working_papers/w33641/w33641.pdf","https://doi.org/10.1287/orsc.2025.20702"],"source_refs":["AIBD-0671-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"AI let individuals match teams.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Individuals without AI.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"0.37 SD (individuals with AI), 0.39 SD (teams with AI), 0.24 SD (teams without AI).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"GenAI tool during one workshop task, after one hour of training.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Blinded expert quality scores, time on task, self-reported emotions.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Standardized solution quality.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Preregistered; data not public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"776 P&G professionals in R&D and commercial roles.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One-day workshop tasks; quality judged as proposals; P&G funding and co-authors.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0671-reported-measure","evaluation_data":null},{"case_id":"AIBD-0025","case_url":"records/AIBD-0025.html","measure":"displayed values such as QT 506 ms and QTc 457 ms","reported_result":"displayed values such as QT 506 ms and QTc 457 ms","baseline":"a QTc 50 to 70 ms longer, as the report states","result":"QTc of 430 ms in repeated measurements on nadolol","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11628798/","https://europepmc.org/article/PMC/PMC11628798","https://pmc.ncbi.nlm.nih.gov/articles/PMC9259864/"],"source_refs":["AIBD-0025-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the watch displayed values such as QT 506 ms and QTc 457 ms","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the watch displayed values such as QT 506 ms and QTc 457 ms","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0025-reported-measure","evaluation_data":null},{"case_id":"AIBD-0007","case_url":"records/AIBD-0007.html","measure":"Crews put the fire out before it spread beyond a 30 x 30 foot area","reported_result":"Crews put the fire out before it spread beyond a 30 x 30 foot area","baseline":"Nobody had called 911","result":"put it out before it spread beyond a 30 x 30 foot area","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.firerescue1.com/artificial-intelligence/over-1-000-ai-cameras-help-spot-calif-wildfires-before-the-first-911-call","https://phys.org/news/2025-09-ai-cameras-wildfires-california-humans.html","https://gulfnews.com/world/americas/california-turns-to-ai-to-help-spot-wildfires-1.1691753082670","https://www.gov.ca.gov/2023/10/24/time-recognizes-cal-fire-ai-alertcalifornia-as-a-best-invention-of-2023/","https://www.ksby.com/santa-margarita/ai-technology-detects-fire-in-san-luis-obispo-county-before-911-calls-are-made","https://inquirer.com/news/nation-world/remote-cameras-ai-spot-wildfires-crew-satellite-wifi-california-20261015.html"],"source_refs":["AIBD-0007-source-0001","AIBD-0007-source-0002","AIBD-0007-source-0003","AIBD-0007-source-0004","AIBD-0007-source-0005","AIBD-0007-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"waiting for a 911 call from the public","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: Nobody had called 911. Result: put it out before it spread beyond a 30 x 30 foot area.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"crews put the fire out before it spread beyond a 30 x 30 foot area","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0007-reported-measure","evaluation_data":null},{"case_id":"AIBD-0162","case_url":"records/AIBD-0162.html","measure":"the early notice kept it to one-eighth of an acre","reported_result":"the early notice kept it to one-eighth of an acre","baseline":"the process of finding, which often time with wildfires can take an hour","result":"get a location, an actual lat/long and drive straight through it","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.cbsnews.com/colorado/news/ai-powered-fire-spotting-cameras-dispatch-firefighters-wildfire/","https://www.kjct8.com/2026/05/05/xcel-energy-deploys-ai-cameras-detect-wildfires-early-colorado/","https://www.summitdaily.com/news/wellington-wildfire-breckenridge-colorado/"],"source_refs":["AIBD-0162-source-0001","AIBD-0162-source-0002","AIBD-0162-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: locating a reported wildfire without the camera's coordinates","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: the process of finding, which often time with wildfires can take an hour. Result: get a location, an actual lat/long and drive straight through it.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the early notice kept it to one-eighth of an acre","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0162-reported-measure","evaluation_data":null},{"case_id":"AIBD-0282","case_url":"records/AIBD-0282.html","measure":"eight new O3 candidates with p_astro > 0.5 (cumulative 5.94)","reported_result":"eight new O3 candidates with p_astro > 0.5 (cumulative 5.94)","baseline":"36 events (all GWTC pipelines); 37 (OGC); 34 (IASa); 36 (IASb)","result":"42","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://arxiv.org/html/2407.07820","https://iopscience.iop.org/article/10.1088/2632-2153/adb5ed","https://arxiv.org/abs/2407.07820","https://arxiv.org/abs/2509.05283","https://arxiv.org/html/2512.04516","https://arxiv.org/html/2512.17204","https://arxiv.org/html/2410.15191v2"],"source_refs":["AIBD-0282-source-0001","AIBD-0282-source-0002","AIBD-0282-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the matched-filtering catalogs (all GWTC pipelines, OGC, IASa and IASb) within AresGW's effective training range","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 36 events (all GWTC pipelines); 37 (OGC); 34 (IASa); 36 (IASb). Result: 42.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"eight new O3 candidates with p_astro > 0.5 (cumulative 5.94)","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0282-reported-measure","evaluation_data":null},{"case_id":"AIBD-0288","case_url":"records/AIBD-0288.html","measure":"recorded a 165 bpm wide-complex tachycardia during palpitations","reported_result":"recorded a 165 bpm wide-complex tachycardia during palpitations","baseline":"VT would have gone undiagnosed","result":"a wide complex tachycardia with a rate of 165 beats/min on his smartwatch","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12245455/","https://web.archive.org/web/2026/https://pubmed.ncbi.nlm.nih.gov/41259841/"],"source_refs":["AIBD-0288-source-0001","AIBD-0288-source-0002","AIBD-0288-source-0003","AIBD-0288-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the clinician's evaluation of palpitations without considering the smartwatch recording","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: VT would have gone undiagnosed. Result: a wide complex tachycardia with a rate of 165 beats/min on his smartwatch.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"recorded a 165 bpm wide-complex tachycardia during palpitations","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0288-reported-measure","evaluation_data":null},{"case_id":"AIBD-0293","case_url":"records/AIBD-0293.html","measure":"24-hour Holter recorded 2,078 ventricular events (1.75% of beats, no couplets or runs)","reported_result":"24-hour Holter recorded 2,078 ventricular events (1.75% of beats, no couplets or runs)","baseline":"Resting heart rate 83 bpm; RMSSD 15.2 ms","result":"Resting heart rate 74 bpm; RMSSD 32.8 ms, two months later","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12268106/"],"source_refs":["AIBD-0293-source-0001","AIBD-0293-source-0002","AIBD-0293-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a 24-hour Holter recorded 2,078 ventricular events (1.75% of beats, no couplets or runs)","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a 24-hour Holter recorded 2,078 ventricular events (1.75% of beats, no couplets or runs)","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0293-reported-measure","evaluation_data":null},{"case_id":"AIBD-0040","case_url":"records/AIBD-0040.html","measure":null,"reported_result":null,"baseline":"The voice sounded more like a robot than a human","result":"it is more me than I ever could have hoped I could hear again","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.pbs.org/newshour/politics/a-neurological-disorder-took-rep-jennifer-wextons-voice-ai-helped-her-bring-it-back-to-the-house-floor","https://www.govinfo.gov/content/pkg/CREC-2024-07-25/html/CREC-2024-07-25-pt1-PgH4934-6.htm","https://history.house.gov/Blog/2026/March/3-27-OralHistory-Wexton/","https://www.washingtonian.com/2024/12/31/rep-jennifer-wextons-way-through/","https://web.archive.org/web/2026/https://www.politico.com/news/magazine/2024/09/09/ai-jennifer-wexton-qa-00177949"],"source_refs":["AIBD-0040-source-0001","AIBD-0040-source-0002","AIBD-0040-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: a traditional text-to-speech app with a generic voice","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: The voice sounded more like a robot than a human. Result: it is more me than I ever could have hoped I could hear again.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0040-reported-measure","evaluation_data":null},{"case_id":"AIBD-0086","case_url":"records/AIBD-0086.html","measure":"translated 36,271 Supreme Court judgments into Hindi and 17,142 into 16 languages","reported_result":"translated 36,271 Supreme Court judgments into Hindi and 17,142 into 16 languages","baseline":"2,238 translated judgments on the portal in January 2023 (and 31 translated from March 2020 to October 2021, per the Supreme Court Observer)","result":"over 31,000 by the end of 2023, and 53,413 (36,271 in Hindi and 17,142 in 16 other languages) by 5 August 2024","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.thehindu.com/news/national/scs-translation-projects-raced-ahead-in-2023-as-retd-hc-judges-law-clerks-help-ai/article67692773.ece","https://www.pib.gov.in/PressReleaseIframePage.aspx?PRID=2043476","https://www.frontiersin.org/journals/political-science/articles/10.3389/fpos.2025.1553705/full","https://cdnbbsr.s3waas.gov.in/s3ec0490f1f4972d133619a60c30f3559e/uploads/2026/01/2026011698.pdf","https://cdnbbsr.s3waas.gov.in/s3ec0490f1f4972d133619a60c30f3559e/uploads/2026/01/2026010287.pdf","https://www.scobserver.in/journal/the-supreme-courts-translation-project-is-slowing-to-a-halt/","https://www.hindustantimes.com/india-news/scs-novel-fix-for-justice-lost-in-translation-101761246271952.html"],"source_refs":["AIBD-0086-source-0002","AIBD-0086-source-0003","AIBD-0086-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"translated 36,271 Supreme Court judgments into Hindi and 17,142 into 16 languages","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"translated 36,271 Supreme Court judgments into Hindi and 17,142 into 16 languages","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0086-reported-measure","evaluation_data":null},{"case_id":"AIBD-0659","case_url":"records/AIBD-0659.html","measure":"Agonist hit rate and potency.","reported_result":"Eighteen AlphaFold compounds were TAAR1 agonists, a hit rate of 60%","baseline":"22% hit rate (7 of 32 compounds)","result":"60% hit rate (18 of 30 compounds)","baseline_value":null,"result_value":null,"unit":null,"effect":"60% against 22% (54% against 22% after adjusting for overlap).","design":"controlled observational.","sample":"30 AlphaFold and 32 homology-model compounds; wild-type and TAAR1 knockout mice.","uncertainty":"One receptor; small numbers tested; animal data only; five authors hold a patent application on the compounds.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11305387/"],"source_refs":["AIBD-0659-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"More than twofold higher hit rate.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Homology-model screen.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"60% against 22% (54% against 22% after adjusting for overlap).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Laboratory screening.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Cell-based TAAR1 activation assays of docked compounds; mouse behavior and temperature tests for one lead.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Agonist hit rate and potency.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Compound data in the paper and supplement.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"30 AlphaFold and 32 homology-model compounds; wild-type and TAAR1 knockout mice.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One receptor; small numbers tested; animal data only; five authors hold a patent application on the compounds.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0659-reported-measure","evaluation_data":null},{"case_id":"AIBD-0087","case_url":"records/AIBD-0087.html","measure":"cut the reference value from R$549.66 million to R$436.47 million","reported_result":"cut the reference value from R$549.66 million to R$436.47 million","baseline":"R$549,659,968.00 (price base October 2023)","result":"R$436,467,943.67 in the final errata of August 2024, a cut of R$113,192,024.33; CGU attributes almost R$100 million of the cut to its findings","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://eaud.cgu.gov.br/relatorios/download/1634659","https://web.archive.org/web/20260207114437id_/https://www.gov.br/cgu/pt-br/assuntos/noticias/2024/12/orgaos-federais-economizaram-em-2024-257-milhoes-de-reais-com-apoio-do-robo-alice-da-controladoria-geral-da-uniao","https://convergenciadigital.com.br/governo/robo-da-cgu-rastreia-r-25-bilhoes-em-licitacoes/","https://web.archive.org/web/20250114172015id_/https://www.gov.br/cgu/pt-br/assuntos/noticias/2024/12/atuacao-do-robo-alice-segue-auxiliando-auditores-a-realizar-trabalhos-preventivos-que-ajudam-a-evitar-desperdicios-e-irregularidades/"],"source_refs":["AIBD-0087-source-0001","AIBD-0087-source-0002","AIBD-0087-source-0003","AIBD-0087-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the tender's original reference value for the engineering services (figure not stated)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: the original reference value, not stated. Result: redução de cerca de R$ 100 milhões no valor de referência.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"reduced the reference value by about R$100 million","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0087-reported-measure","evaluation_data":null},{"case_id":"AIBD-0095","case_url":"records/AIBD-0095.html","measure":"weekly incomes rising from ₦4,500 to ₦10,000 and food shortages falling from 97% to 7%","reported_result":"weekly incomes rising from ₦4,500 to ₦10,000 and food shortages falling from 97% to 7%","baseline":"weekly income ₦4,500; 97% facing food shortages","result":"weekly income ₦10,000; 7% facing food shortages","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.givedirectly.org/flood-forecast-ai/","https://www.thenewhumanitarian.org/investigations/2026/08/31/givedirectlys-ai-experiment-left-flood-victims-behind-nigeria","https://www.aljazeera.com/features/2024/9/21/could-ai-save-nigerians-from-devastating-floods","https://restofworld.org/2025/google-flood-hub-cash-aid/"],"source_refs":["AIBD-0095-source-0001","AIBD-0095-source-0002","AIBD-0095-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"recipients' own situation before the transfers (GiveDirectly follow-up survey; no control group)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: weekly income ₦4,500; 97% facing food shortages. Result: weekly income ₦10,000; 7% facing food shortages.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"weekly incomes rising from ₦4,500 to ₦10,000 and food shortages falling from 97% to 7%","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0095-reported-measure","evaluation_data":null},{"case_id":"AIBD-0647","case_url":"records/AIBD-0647.html","measure":"Clinic attendance within 8 weeks.","reported_result":"raised clinic attendance for referable retinopathy from 18.2% to 35.9%","baseline":"18.2% of people with referable retinopathy attended the eye clinic within 8 weeks","result":"35.9% attended within 8 weeks","baseline_value":null,"result_value":null,"unit":null,"effect":"35.9% vs 18.2%; odds ratio 2.52.","design":"randomized.","sample":"2,378 adults with diabetes at 24 hospitals in Kilimanjaro and Arusha, 8 March 2023 to 31 August 2024; 904 with referable disease.","uncertainty":"AI effect not separated from face-to-face counselling; attendance remained low in both groups.","comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/42863478/","https://pmc.ncbi.nlm.nih.gov/articles/PMC10824006/"],"source_refs":["AIBD-0647-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"17.7 percentage points.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Deferred human grading and telephone counselling.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"35.9% vs 18.2%; odds ratio 2.52.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"AI grading with immediate result for all participants in the intervention group.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Attendance at the central ophthalmology clinic within 8 weeks, among participants with referable retinopathy by masked UK-certified graders.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Clinic attendance within 8 weeks.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Registered as ISRCTN18317152; protocol published in BMJ Open.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"2,378 adults with diabetes at 24 hospitals in Kilimanjaro and Arusha, 8 March 2023 to 31 August 2024; 904 with referable disease.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"AI effect not separated from face-to-face counselling; attendance remained low in both groups.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0647-reported-measure","evaluation_data":null},{"case_id":"AIBD-0173","case_url":"records/AIBD-0173.html","measure":"$1 billion was recovered through machine-learning identification of Treasury check fraud","reported_result":"$1 billion was recovered through machine-learning identification of Treasury check fraud","baseline":"$652.7 million in FY23","result":"over $4 billion in fraud and improper payments this fiscal year (FY) (October 2023 – September 2024)","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://home.treasury.gov/news/press-releases/jy2650","https://www.nextgov.com/artificial-intelligence/2024/10/ai-tools-helped-treasury-recover-billions-fraud-and-improper-payments/400368/","https://www.nbcnews.com/business/consumer/how-ai-artificial-intelligence-fights-taxpayer-fraud-treasury-dept-rcna175916","https://fiscal.treasury.gov/about-us/news/treasurys-bureau-fiscal-service-recognized-for-payment-integrity-efforts","https://www.cnn.com/2024/10/17/business/ai-fraud-treasury/index.html"],"source_refs":["AIBD-0173-source-0001","AIBD-0173-source-0002","AIBD-0173-source-0003","AIBD-0173-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Treasury's fraud and improper-payment prevention and recovery in FY2023, before the enhanced processes","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: $652.7 million in FY23. Result: over $4 billion in fraud and improper payments this fiscal year (FY) (October 2023 – September 2024).","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"$1 billion recovered through machine-learning check-fraud detection","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0173-reported-measure","evaluation_data":null},{"case_id":"AIBD-0380","case_url":"records/AIBD-0380.html","measure":"traced three tiers of card transfers and froze 500,000 yuan within 10 minutes","reported_result":"traced three tiers of card transfers and froze 500,000 yuan within 10 minutes","baseline":"A team of five or six officers working for two weeks before the first arrests","result":"Nine arrests on the third day, with 500,000 yuan frozen in 10 minutes","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.suzhou.gov.cn/2024ndzt/mtjjxzc/202410/41e0e0bc1a8843ebb55e2a0d54f44bf5.shtml","https://m.163.com/dy/article/JDQORBU7053469KC.html","https://oecd.ai/en/incidents/2024-10-03-edf0"],"source_refs":["AIBD-0380-source-0002","AIBD-0380-source-0001","AIBD-0380-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"traced three tiers of card transfers and froze 500,000 yuan within 10 minutes","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"traced three tiers of card transfers and froze 500,000 yuan within 10 minutes","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0380-reported-measure","evaluation_data":null},{"case_id":"AIBD-0286","case_url":"records/AIBD-0286.html","measure":"radiofrequency ablation restored his ejection fraction to 50% at six months","reported_result":"radiofrequency ablation restored his ejection fraction to 50% at six months","baseline":"15%","result":"50%","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12209856/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12786555/"],"source_refs":["AIBD-0286-source-0001","AIBD-0286-source-0002","AIBD-0286-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the same patient's LVEF before the radiofrequency ablation that the smartwatch alert led to (a treatment effect, not a with-versus-without-device comparison)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 15%. Result: 50%.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"radiofrequency ablation restored LVEF to 50% at six months","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0286-reported-measure","evaluation_data":null},{"case_id":"AIBD-0197","case_url":"records/AIBD-0197.html","measure":"Correlation between the daily AI respiratory health status and the human clinical coughing score; association of pathogen loads with the score.","reported_result":"Spearman's rho of -0.478 (batch 1) and -0.468 (batch 2) between the AI score and the human coughing score; high swine influenza A loads were associated with a lower score.","baseline":"twenty human clinical coughing scores per batch, with a coughing index of at most 0.4% (batch 1) and a maximum of 1.26% (batch 2)","result":"a 24/7 AI respiratory health score that fell as the human-scored coughing rose (Spearman's rho −0.478 and −0.468)","baseline_value":null,"result_value":null,"unit":null,"effect":"Spearman's rho of -0.478 (batch 1) and -0.468 (batch 2) between the AI score and the human coughing score; high swine influenza A loads were associated with a lower score.","design":"case study.","sample":"One conventional piglet-producing farm in Germany; a nursery designed for 600 pigs per airspace; two batches, each monitored for 39 days.","uncertainty":"One farm and two batches; association only, with no alert acted on; Boehringer Ingelheim funded the study and three of its employees are co-authors.","comparison_basis":"measured","source_urls":["https://europepmc.org/article/PMC/PMC11512249","https://doi.org/10.3390/v16101575","https://doi.org/10.3390/v18010072","https://europepmc.org/article/PMC/PMC11662573"],"source_refs":["https://europepmc.org/article/PMC/PMC11512249"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Human clinical coughing scores taken 20 times per batch.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Spearman's rho of -0.478 (batch 1) and -0.468 (batch 2) between the AI score and the human coughing score; high swine influenza A loads were associated with a lower score.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two SoundTalks monitors counted coughs around the clock in a blinded mode; veterinarians scored coughing by hand and screened oral fluids and air samples by qPCR.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Correlation between the daily AI respiratory health status and the human clinical coughing score; association of pathogen loads with the score.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"One conventional piglet-producing farm in Germany; a nursery designed for 600 pigs per airspace; two batches, each monitored for 39 days.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One farm and two batches; association only, with no alert acted on; Boehringer Ingelheim funded the study and three of its employees are co-authors.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0197-reported-measure","evaluation_data":null},{"case_id":"AIBD-0091","case_url":"records/AIBD-0091.html","measure":"no further emissions have been detected at the location since 15 October 2024","reported_result":"no further emissions have been detected at the location since 15 October 2024","baseline":"methane plumes of about 2,100 to 3,600 kilograms per hour on 17 EMIT overpasses from 17 August 2022 to 29 September 2024 (Carbon Mapper); 113 plumes since 1 July 1987 and an estimated 960,000 tonnes in total (UNEP)","result":"no emissions above satellite detection limits at the location since 15 October 2024 (UNEP); no further plume at that point in Carbon Mapper's catalog","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/20261004073839id_/https://wedocs.unep.org/rest/api/core/bitstreams/5810b829-cf15-4eb0-b509-8dd489884a66/content","https://wedocs.unep.org/bitstreams/5810b829-cf15-4eb0-b509-8dd489884a66/download","https://wedocs.unep.org/items/33a43a3a-8a2a-4831-8f18-e09e472623b0","https://news.un.org/en/story/2026/07/1168007","https://www.unep.org/topics/energy/methane/methane-alert-and-response-system-mars","https://www.unep.org/en/what-we-do/energy/methane/international-methane-emissions-observatory/mars-frequently-asked","https://methanedata.unep.org/downloads/top50_emitters/latest/mars_snapshot_latest.pdf","https://wedocs.unep.org/bitstream/handle/20.500.11822/48664/An-Eye-on-Methane-2025.pdf?sequence=1&isAllowed=y","https://api.carbonmapper.org/api/v1/catalog/plumes/annotated?bbox=5.985&bbox=31.770&bbox=6.005&bbox=31.786&limit=100"],"source_refs":["AIBD-0091-source-0001","AIBD-0091-source-0002","AIBD-0091-source-0003","AIBD-0091-source-0004","AIBD-0091-source-0005","AIBD-0091-source-0006","AIBD-0091-source-0007"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0091-reported-measure","evaluation_data":null},{"case_id":"AIBD-0468","case_url":"records/AIBD-0468.html","measure":"enabled Erin Taylor, who has ALS, to introduce Lenovo's CEO at Tech World","reported_result":"enabled Erin Taylor, who has ALS, to introduce Lenovo's CEO at Tech World","baseline":null,"result":"Spoke on stage through an avatar in a recreation of her pre-diagnosis voice","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://news.lenovo.com/pressroom/press-releases/ai-powered-communication-solution-for-people-with-als-revealed-tech-world/","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-519","https://alsnetwork.org/spotlight-on-erin-taylor/","https://observer.com/2025/12/built-by-and-for-people-with-paralysis-this-als-tech-gives-a-i-a-human-voice/"],"source_refs":["AIBD-0468-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0468-reported-measure","evaluation_data":null},{"case_id":"AIBD-0009","case_url":"records/AIBD-0009.html","measure":"saved 414 elephants in 2023 and 383 up to 16 October 2024","reported_result":"saved 414 elephants in 2023 and 383 up to 16 October 2024","baseline":"on an average 20 elephants die due to train collisions in the country every year and a majority of these incidents take place in the Northeast Frontier Railway","result":"since the launch of the system, these 11 corridors have not reported any train-elephant collision","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.tribuneindia.com/news/india/ai-based-surveillance-mechanism-helps-eliminate-elephant-deaths-due-to-train-collisions-in-11-northeast-corridors-541918","https://morungexpress.com/alert-train-drivers-avert-collision-with-herd-of-elephants-in-assam","https://www.sentinelassam.com/breakingnews/assam-ai-technology-generated-alerts-prevent-train-from-colliding-with-elephants","https://themeghalayanexpress.com/ai-technology-alerts-drivers-to-stop-trains-to-rescue-a-herd-of-elephants-in-assam/","https://ddnews.gov.in/en/indian-railways-expands-ai-based-intrusion-detection-to-prevent-elephant-collisions-across-key-routes/","https://www.sentinelassam.com/north-east-india-news/assam-news/ai-shield-and-plan-bee-save-255-elephants-across-northeast-frontier-railway","https://timesofindia.indiatimes.com/city/guwahati/intrusion-detection-system-could-have-averted-tragedy/articleshow/126097369.cms"],"source_refs":["AIBD-0009-source-0001","AIBD-0009-source-0002","AIBD-0009-source-0003","AIBD-0009-source-0004","AIBD-0009-source-0005","AIBD-0009-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"train-elephant collisions before the system (national average of elephant deaths on tracks)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: on an average 20 elephants die due to train collisions in the country every year and a majority of these incidents take place in the Northeast Frontier Railway. Result: since the launch of the system, these 11 corridors have not reported any train-elephant collision.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"saved 414 elephants in 2023 and 383 up to 16 October 2024","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0009-reported-measure","evaluation_data":null},{"case_id":"AIBD-0065","case_url":"records/AIBD-0065.html","measure":"New HCM diagnoses among flagged patients; sensitivity, specificity and positive predictive value at several probability thresholds.","reported_result":"by week 8 the resting gradient was 9 mm Hg","baseline":"interpreted clinically as normal sinus rhythm with first-degree atrioventricular block","result":"a high probability score (>0.95) for HCM","baseline_value":null,"result_value":null,"unit":null,"effect":"1,265 patients flagged; 511 had known HCM, 63 had a new HCM diagnosis and 691 had another diagnosis; positive predictive value 66% at the 0.95 threshold, with 50% sensitivity.","design":"other.","sample":"Cleveland Clinic, Cleveland, Ohio; 103,492 ECGs in 45,873 patients, 19 February 2024 to 1 November 2024.","uncertainty":"Single center; authors include two consultants for the developer; the abstract reports no patient outcomes after diagnosis.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12046996/","https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN230003.pdf","https://pubmed.ncbi.nlm.nih.gov/40100197/"],"source_refs":["AIBD-0065-source-0003"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Prior clinical evaluation and confirmed diagnosis in the medical record.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"1,265 patients flagged; 511 had known HCM, 63 had a new HCM diagnosis and 691 had another diagnosis; positive predictive value 66% at the 0.95 threshold, with 50% sensitivity.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Viz HCM run on every 12-lead ECG.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Prospective analysis of all hospital ECGs with chart review of each flagged patient for confirmed, new or alternative diagnoses.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"New HCM diagnoses among flagged patients; sensitivity, specificity and positive predictive value at several probability thresholds.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Cleveland Clinic, Cleveland, Ohio; 103,492 ECGs in 45,873 patients, 19 February 2024 to 1 November 2024.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Single center; authors include two consultants for the developer; the abstract reports no patient outcomes after diagnosis.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0065-reported-measure","evaluation_data":null},{"case_id":"AIBD-0668","case_url":"records/AIBD-0668.html","measure":"Quality grades and productivity.","reported_result":"raised productivity on five of six tasks by 50% to 130%","baseline":"Productivity and graded quality without AI","result":"Higher quality; productivity gains of 50% to 130% on five of six tasks","baseline_value":null,"result_value":null,"unit":null,"effect":"Significant quality gains for both tools; productivity gains of 50% to 130% on five of six tasks.","design":"randomized.","sample":"153 enrolled upper-level law students at Michigan and Minnesota; 137 completed at least one task, 125 all six.","uncertainty":"Students in place of practicing lawyers; designed tasks; graders were co-authors; o1-preview work had more hallucinated citations.","comparison_basis":"measured","source_urls":["https://web.archive.org/web/2026id_/https://scholarship.law.umn.edu/cgi/viewcontent.cgi?article=2220&context=faculty_articles","https://web.archive.org/web/2025/https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5162111"],"source_refs":["AIBD-0668-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Productivity up 50% to 130% on five tasks.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"No AI.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Significant quality gains for both tools; productivity gains of 50% to 130% on five of six tasks.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Vincent AI or o1-preview during the tasks.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Graded quality and productivity on six legal tasks.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Quality grades and productivity.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Task texts in Appendix B of the article.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"153 enrolled upper-level law students at Michigan and Minnesota; 137 completed at least one task, 125 all six.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Students in place of practicing lawyers; designed tasks; graders were co-authors; o1-preview work had more hallucinated citations.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0668-reported-measure","evaluation_data":null},{"case_id":"AIBD-0124","case_url":"records/AIBD-0124.html","measure":null,"reported_result":null,"baseline":"manual code analysis and hand-written fuzz tests","result":"one command, after which the agent analyzed the code and generated and ran the test","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.prnewswire.com/news-releases/code-intelligence-launches-spark-its-ai-test-agent-that-autonomously-uncovers-vulnerabilities-without-human-interaction-302345548.html","https://github.com/wolfSSL/wolfssl/pull/8139","https://github.com/wolfSSL/wolfssl/releases/tag/v5.7.6-stable","https://web.archive.org/web/20250109141038/https://www.code-intelligence.com/blog/ai-generated-fuzz-test-wolfssl-vulnerability","https://qa-financial.com/german-startup-to-launch-worlds-first-ai-test-agent/","https://www.scworld.com/feature/how-ai-can-revolutionize-vulnerability-research"],"source_refs":["AIBD-0124-source-0002","AIBD-0124-source-0001","AIBD-0124-source-0003","AIBD-0124-source-0004","AIBD-0124-source-0005","AIBD-0124-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0124-reported-measure","evaluation_data":null},{"case_id":"AIBD-0345","case_url":"records/AIBD-0345.html","measure":"eight methane leaks in Turkmenistan that were mitigated from November 2024 to March 2026","reported_result":"eight methane leaks in Turkmenistan that were mitigated from November 2024 to March 2026","baseline":"47 plumes detected from 22 June 2025, with an estimated 16,000 tonnes of methane released","result":"No emissions above the satellites' detection limits after 7 March 2026","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/20260916184457id_/https://wedocs.unep.org/rest/api/core/bitstreams/c844f45e-7bcb-4efc-a636-800724751ba5/content","https://www.unep.org/en/what-we-do/energy/methane/mars-case-studies","https://www.theguardian.com/environment/2026/aug/31/turkmenistan-methane-leaks-un-alerts-mars","https://www.unep.org/en/what-we-do/energy/methane/methane-alert-and-response-system-mars","https://wedocs.unep.org/bitstreams/c844f45e-7bcb-4efc-a636-800724751ba5/download"],"source_refs":["AIBD-0345-source-0001","AIBD-0345-source-0002","AIBD-0345-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"eight listed mitigation cases; 192 alerts and about 20% answered in the August 2026 report","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"eight listed mitigation cases; 192 alerts and about 20% answered in the August 2026 report","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0345-reported-measure","evaluation_data":null},{"case_id":"AIBD-0620","case_url":"records/AIBD-0620.html","measure":"Operational performance score.","reported_result":"The ChatGPT group scored 73.12 against 65.54 for the video-only group (P < .001)","baseline":"Mean skills score 65.54 (videos only)","result":"Mean skills score 73.12 (videos with ChatGPT)","baseline_value":null,"result_value":null,"unit":null,"effect":"73.12 against 65.54 (P < .001).","design":"randomized.","sample":"187 undergraduate dental students at a Chinese university, October to November 2024.","uncertainty":"One-week intervention; comparison group had no added support; accuracy of ChatGPT answers not checked; simulated test.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12152432/"],"source_refs":["AIBD-0620-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The gain is attributed to adding ChatGPT to video learning.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Video-only learning.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"73.12 against 65.54 (P < .001).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One week of skill learning.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Operational skill score on a desktop virtual reality simulator; pupil diameter for cognitive load; questionnaires.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Operational performance score.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"ClinicalTrials.gov NCT06942130.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"187 undergraduate dental students at a Chinese university, October to November 2024.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One-week intervention; comparison group had no added support; accuracy of ChatGPT answers not checked; simulated test.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0620-reported-measure","evaluation_data":null},{"case_id":"AIBD-0270","case_url":"records/AIBD-0270.html","measure":null,"reported_result":null,"baseline":"Human experts' conclusions on the published MRI findings of cerebral intravascular large B-cell lymphoma","result":"ChatGPT-4o judged the common MRI findings useful for recognition but not specific enough for diagnosis, and the authors state that this agrees with the experts","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11602322/"],"source_refs":["AIBD-0270-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0270-reported-measure","evaluation_data":null},{"case_id":"AIBD-0056","case_url":"records/AIBD-0056.html","measure":null,"reported_result":null,"baseline":"She was having trouble reading printed materials, specifically recipes and food preparation instructions, through her habitual spectacle lenses","result":"Seeing AI helping her to access printed recipes","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12417062/","https://apps.apple.com/us/app/seeing-ai/id999062298"],"source_refs":["AIBD-0056-source-0001","AIBD-0056-source-0002","AIBD-0056-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: reading print through her habitual spectacle lenses","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: She was having trouble reading printed materials, specifically recipes and food preparation instructions, through her habitual spectacle lenses. Result: Seeing AI helping her to access printed recipes.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0056-reported-measure","evaluation_data":null},{"case_id":"AIBD-0047","case_url":"records/AIBD-0047.html","measure":"Dose-limiting toxicities, adverse reactions, half-life and serum live-virus neutralization.","reported_result":"dose-dependent live-virus neutralization at 600 and 1,200 mg","baseline":null,"result":"Dose-dependent increases in serum live virus neutralization at 600 and 1,200 mg, separated from placebo","baseline_value":null,"result_value":null,"unit":null,"effect":"No dose-limiting toxicities, adverse reactions of grade 1 or 2 only, a half-life of 54 days, and dose-dependent neutralization at 600 and 1,200 mg separated from placebo.","design":"randomized.","sample":"51 healthy adults aged 18 to 55 at one phase 1 unit in Orlando, Florida; 36 received GB-0669 and 15 placebo across five single intravenous dose cohorts from 100 to 2,400 mg.","uncertainty":"Healthy volunteers only, small cohorts, serum neutralization in the laboratory with no infection or clinical endpoint; the authors are company staff, and the registry lists the allocation as non-randomized.","comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/42446369/","https://doi.org/10.1093/infdis/jiag349","https://www.biorxiv.org/content/10.1101/2025.03.30.646023v1","https://clinicaltrials.gov/study/NCT07050511","https://pmc.ncbi.nlm.nih.gov/articles/PMC11777668/"],"source_refs":["https://pubmed.ncbi.nlm.nih.gov/42446369/"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not measured; the trial does not compare the AI-optimized antibody with one found by another method.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Placebo.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"No dose-limiting toxicities, adverse reactions of grade 1 or 2 only, a half-life of 54 days, and dose-dependent neutralization at 600 and 1,200 mg separated from placebo.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One intravenous dose.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Safety monitoring, serum drug concentrations and live-virus neutralization assays of participants' serum over 43 weeks.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Dose-limiting toxicities, adverse reactions, half-life and serum live-virus neutralization.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"51 healthy adults aged 18 to 55 at one phase 1 unit in Orlando, Florida; 36 received GB-0669 and 15 placebo across five single intravenous dose cohorts from 100 to 2,400 mg.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Healthy volunteers only, small cohorts, serum neutralization in the laboratory with no infection or clinical endpoint; the authors are company staff, and the registry lists the allocation as non-randomized.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0047-reported-measure","evaluation_data":null},{"case_id":"AIBD-0477","case_url":"records/AIBD-0477.html","measure":null,"reported_result":null,"baseline":null,"result":"work efficiency more than ten times higher in design work, by the developer's preliminary estimate","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.ainnovation.com/art/detail/474/","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-574"],"source_refs":["AIBD-0477-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0477-reported-measure","evaluation_data":null},{"case_id":"AIBD-0486","case_url":"records/AIBD-0486.html","measure":null,"reported_result":null,"baseline":"manual negotiation by merchandise staff, with the time per negotiation not stated","result":"negotiation time reduced by 80 percent and average margins 5 to 6 percent higher, by the developer's account","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://tlo.korea.ac.kr/en/partnership/1103","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-629","https://news.jkn.co.kr/post/852656","https://magazine.hankyung.com/job-joy/article/202602057583d"],"source_refs":["AIBD-0486-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0486-reported-measure","evaluation_data":null},{"case_id":"AIBD-0640","case_url":"records/AIBD-0640.html","measure":"Share meeting the composite outcome; program initiation and completion.","reported_result":"31.7% of the AI group and 31.9% of the human-coaching group met the combined goal","baseline":"82.7% started the human-led program","result":"93.4% started the AI-led program","baseline_value":null,"result_value":null,"unit":null,"effect":"31.7% against 31.9% (noninferior); initiation 93.4% against 82.7%; completion 63.9% against 50.3%.","design":"randomized.","sample":"368 adults with prediabetes and overweight or obesity at two US sites.","uncertainty":"Unmasked; surrogate outcomes; human programs delivered remotely; app updates during the trial.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12560030"],"source_refs":["AIBD-0640-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Equal outcome to human coaching with higher uptake.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Human-led CDC-recognized diabetes prevention programs.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"31.7% against 31.9% (noninferior); initiation 93.4% against 82.7%; completion 63.9% against 50.3%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"12 months.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Composite of at least 5% weight loss, at least 4% weight loss with 150 minutes of weekly activity by actigraphy, or HbA1c drop of 0.2 points at 12 months.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Share meeting the composite outcome; program initiation and completion.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"ClinicalTrials.gov NCT05056376; protocol in supplements.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"368 adults with prediabetes and overweight or obesity at two US sites.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Unmasked; surrogate outcomes; human programs delivered remotely; app updates during the trial.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0640-reported-measure","evaluation_data":null},{"case_id":"AIBD-0558","case_url":"records/AIBD-0558.html","measure":null,"reported_result":null,"baseline":"Advice from manual searching or a phone call to the specialist","result":"About 15 minutes saved per conversation with the assistant","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.datnexa.com/post/what-peterborough-s-asc-team-learned-from-deploying-an-ai-assistant","https://www.local.gov.uk/case-studies/peterborough-city-council-hey-geraldine-personalised-ai-assistant","https://www.bbc.com/news/articles/cjelegxjq3do","https://www.peoplemanagement.co.uk/article/1927796/hey-geraldine-council-turns-veteran-worker-ai-chatbot-%E2%80%93-ethical-digitally-clone-staff"],"source_refs":["AIBD-0558-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0558-reported-measure","evaluation_data":null},{"case_id":"AIBD-0140","case_url":"records/AIBD-0140.html","measure":null,"reported_result":null,"baseline":"Both  the doctor who saw her and the radiologist declared her free of breaks and, after answering some of our questions, sent her home with an ace wrap and ibuprofen.","result":"the wrist specialist took new x-rays (multiple views), examined her, and confirmed Grok’s diagnosis of a distal radial head fracture with dorsal displacement.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://x.com/AJKayWriter/status/1878084198205907173","https://www.ndtv.com/feature/woman-claims-elon-musks-grok-detected-her-daughter-s-fracture-after-doctors-fail-to-diagnose-it-7456337","https://www.hindustantimes.com/trending/mother-claims-elon-musk-s-ai-detected-daughter-s-fracture-that-doctors-missed-this-was-eye-opening-for-me-101736662763447.html","https://www.livemint.com/ai/artificial-intelligence/ai-chat-helps-avoid-surgery-elon-musk-s-grok-identifies-teenage-girl-s-wrist-fracture-after-doctors-fail-to-diagnose-it-11736671754200.html","https://www.indiatoday.in/trending-news/story/elon-musk-grok-diagnoses-fracture-missed-by-doctors-us-woman-claims-in-viral-post-2663680-2025-01-12","https://www.livemint.com/technology/tech-news/hey-liar-elon-musk-claims-grok-ai-can-diagnose-medical-injuries-liverdoc-abby-philips-gives-him-a-reality-check-11736678451620.html","https://x.com/elonmusk/status/1878203372513136854","https://timesofindia.indiatimes.com/life-style/health-fitness/health-news/elon-musks-grok-ai-diagnoses-fracture-missed-by-doctors-claims-woman/articleshow/117255256.cms","https://pmc.ncbi.nlm.nih.gov/articles/PMC13115499/","https://pubmed.ncbi.nlm.nih.gov/41785426/"],"source_refs":["AIBD-0140-source-0001","AIBD-0140-source-0002","AIBD-0140-source-0003","AIBD-0140-source-0004","AIBD-0140-source-0005","AIBD-0140-source-0006","AIBD-0140-source-0007","AIBD-0140-source-0008"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the urgent-care doctor's and radiologist's reading of the same X-rays","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: Both  the doctor who saw her and the radiologist declared her free of breaks and, after answering some of our questions, sent her home with an ace wrap and ibuprofen. Result: the wrist specialist took new x-rays (multiple views), examined her, and confirmed Grok’s diagnosis of a distal radial head fracture with dorsal displacement.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0140-reported-measure","evaluation_data":null},{"case_id":"AIBD-0150","case_url":"records/AIBD-0150.html","measure":"avoided roughly $55,000 in penalties and over $18,000 in overdue rent","reported_result":"avoided roughly $55,000 in penalties and over $18,000 in overdue rent","baseline":"she worked with a lawyer from a local tenant advocacy network but lost in a jury trial","result":"White managed to overturn her eviction notice and avoid roughly $55,000 in penalties and more than $18,000 in overdue rent","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.nbcnews.com/tech/innovation/ai-chatgpt-court-law-legal-lawyer-self-represent-pro-se-attorney-rcna230401","https://www.news247.gr/kosmos/ginaika-kerdise-dikastiki-ipothesi-me-dikigoro-to-chatgpt/","https://www.huffingtonpost.gr/diethnes/gynaika-stin-kalifornia-kerdise-dikastiki-ypothesi-me-synigoro-tochatgpt/","https://x.com/ihtesham2005/status/2040399092111864187"],"source_refs":["AIBD-0150-source-0001","AIBD-0150-source-0002","AIBD-0150-source-0003","AIBD-0150-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: representation by a lawyer from a local tenant advocacy network, with whom she lost the jury trial","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: she worked with a lawyer from a local tenant advocacy network but lost in a jury trial. Result: White managed to overturn her eviction notice and avoid roughly $55,000 in penalties and more than $18,000 in overdue rent.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"avoided roughly $55,000 in penalties and over $18,000 in overdue rent","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0150-reported-measure","evaluation_data":null},{"case_id":"AIBD-0202","case_url":"records/AIBD-0202.html","measure":"Shipping age, share shipped at grade 1+, income per head; sow accident rate, piglets shipped per sow per year.","reported_result":"Income per head rose from 100,000 to 118,000 won, up 18%","baseline":"출하일령 182.3일; 1+등급 출하율 35.3%; 두당 소득 10만 원","result":"출하일령 176.7일 (5.6일 단축); 1+등급 출하율 37.7% (2.4%p 증가); 두당 소득 11만 8000원 (18% 증가)","baseline_value":null,"result_value":null,"unit":null,"effect":"Shipping age 182.3 to 176.7 days; grade 1+ share 35.3% to 37.7%; income per head 100,000 to 118,000 won; sow accidents 3.1% to 2.7%; piglets per sow per year 18.3 to 19.2.","design":"before/after.","sample":"Six pig farms in Anseong, Pocheon and Yeoju (weighing system); farms in Pocheon (sow system), 2025.","uncertainty":"No control farms, no statistical test and no published data. The agency that ran the pilot reports the results, and the vendor is not named.","comparison_basis":"measured","source_urls":["https://news.tf.co.kr/read/national/2299220.htm","https://www.joongboo.com/news/articleView.html?idxno=363719065","https://oecd.ai/en/incidents/2026-03-05-1ac3","https://gnews.gg.go.kr/briefing/brief_gongbo_view.do?BS_CODE=S017&number=69516"],"source_refs":["AIBD-0202-source-0004"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The same farms before the systems were introduced.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Shipping age 182.3 to 176.7 days; grade 1+ share 35.3% to 37.7%; income per head 100,000 to 118,000 won; sow accidents 3.1% to 2.7%; piglets per sow per year 18.3 to 19.2.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"The agency's comparison of farm figures before and after the systems were introduced; the method, period and data are not published.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Shipping age, share shipped at grade 1+, income per head; sow accident rate, piglets shipped per sow per year.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Six pig farms in Anseong, Pocheon and Yeoju (weighing system); farms in Pocheon (sow system), 2025.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"No control farms, no statistical test and no published data. The agency that ran the pilot reports the results, and the vendor is not named.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0202-reported-measure","evaluation_data":null},{"case_id":"AIBD-0211","case_url":"records/AIBD-0211.html","measure":"Diagnostic yield of confirmed local diagnoses.","reported_result":"18 confirmed local diagnoses, or 4.8%","baseline":"376 cases unresolved in the local research record, including some diagnosed elsewhere","result":"18 confirmed local diagnoses: 11 new and seven rediscoveries; 4.8% overall","baseline_value":null,"result_value":null,"unit":null,"effect":"18 of 376 cases (4.8%) received confirmed local diagnoses; seven of these 18 were rediscoveries of diagnoses already established elsewhere, which leaves 11 new diagnoses.","design":"before/after.","sample":"376 unsolved cases at Boston Children's Hospital in four cohorts: 100 neurodevelopmental, 61 neuromuscular, 200 sudden unexpected death in pediatrics and 15 early psychosis.","uncertainty":"Retrospective design, small absolute numbers, cohort ascertainment that differed, inputs of variable completeness, and four OpenAI authors including the first author.","comparison_basis":"measured","source_urls":["https://doi.org/10.1056/AIcs2501343","https://www.nbcnews.com/tech/innovation/ai-boston-childrens-hospital-diagnose-rare-diseases-kids-openai-rcna350387"],"source_refs":["https://pmc.ncbi.nlm.nih.gov/articles/PMC13417654/"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not measured; the study has no arm in which experts reanalyzed the same cases with current knowledge.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The prior local research record, with no concurrent human reanalysis.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"18 of 376 cases (4.8%) received confirmed local diagnoses; seven of these 18 were rediscoveries of diagnoses already established elsewhere, which leaves 11 new diagnoses.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Model candidates adjudicated by at least two reviewers under ACMG/AMP criteria; a diagnosis counted only when a CLIA-certified laboratory confirmed a pathogenic or likely pathogenic variant and the result was returned to the family.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Diagnostic yield of confirmed local diagnoses.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"376 unsolved cases at Boston Children's Hospital in four cohorts: 100 neurodevelopmental, 61 neuromuscular, 200 sudden unexpected death in pediatrics and 15 early psychosis.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective design, small absolute numbers, cohort ascertainment that differed, inputs of variable completeness, and four OpenAI authors including the first author.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0211-reported-measure","evaluation_data":null},{"case_id":"AIBD-0251","case_url":"records/AIBD-0251.html","measure":"high-severity close encounters per vessel per month.","reported_result":"a 52% reduction in high-severity close encounters","baseline":"2.7 per vessel per month in the early post-implementation phase","result":"1.3 by months 10–12 - a 52% reduction","baseline_value":null,"result_value":null,"unit":null,"effect":"from about 2.7 to about 1.3 per vessel per month by months 10 to 12, a 52% reduction; 36% in the North and Baltic Seas and 18% in the China and Japan Seas.","design":"before/after.","sample":"139 vessels of mixed types in global trade, 139 ship-years and more than 10.8 million nautical miles, data from 2023 to 2025.","uncertainty":"No control fleet and no confidence intervals. The baseline is the first three months with the system; no period without the system is measured. Crew familiarization, season and route changes are not separated from the system's effect. The vendor measures the outcome with its own sensors, and the co-author insurer subsidizes the system.","comparison_basis":"measured","source_urls":["https://www.orca-ai.io/reports/safetyreport/","https://d3cpegos94401u.cloudfront.net/publications/documents/orca-ai-&-northstandard-safety-report---final.pdf","https://www.rivieramm.com/news-content-hub/news-content-hub/northstandard-study-demonstrates-benefits-of-ai-navigation-support-88473","https://www.hellenicshippingnews.com/joint-orca-ai-northstandard-study-highlights-strong-impact-of-ai-on-navigational-safety/","https://www.marinelink.com/news/study-orca-ai-northstandard-highlights-538079","https://oecd.ai/en/incidents/2026-04-15-b7a4","https://maritime-executive.com/article/top-insurer-subsidizes-use-of-ai-navigation-safety-tool"],"source_refs":["AIBD-0251-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"the same vessels in months 1 to 3 after installation, with no untreated fleet.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"from about 2.7 to about 1.3 per vessel per month by months 10 to 12, a 52% reduction; 36% in the North and Baltic Seas and 18% in the China and Japan Seas.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"each vessel's first 12 months with the SeaPod installed.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Data from Orca AI's own platform on each vessel. Events were classed as high-severity by CPA, TCPA, COLREGs scenario, local traffic density and counterpart vessel behavior. Rates were normalized per vessel per month.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"high-severity close encounters per vessel per month.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"139 vessels of mixed types in global trade, 139 ship-years and more than 10.8 million nautical miles, data from 2023 to 2025.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"No control fleet and no confidence intervals. The baseline is the first three months with the system; no period without the system is measured. Crew familiarization, season and route changes are not separated from the system's effect. The vendor measures the outcome with its own sensors, and the co-author insurer subsidizes the system.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0251-reported-measure","evaluation_data":null},{"case_id":"AIBD-0273","case_url":"records/AIBD-0273.html","measure":"Heart Failure Likelihood Score, exacerbation risk and Forrester classification, with narrative management recommendations.","reported_result":"The system classified one patient as Warm/Wet and recommended maintaining diuretic therapy with intensified monitoring, and classified the other as Cold/Dry and recommended diuretic de-escalation; neither patient needed hospitalization during follow-up.","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":"The system classified one patient as Warm/Wet and recommended maintaining diuretic therapy with intensified monitoring, and classified the other as Cold/Dry and recommended diuretic de-escalation; neither patient needed hospitalization during follow-up.","design":"case study.","sample":"Two heart-failure patients in a skilled nursing facility, chosen to show contrasting Forrester hemodynamic quadrants.","uncertainty":"Two purposively selected cases with no accuracy metric and no comparison group; the senior author founded the platform; the authors call for prospective validation.","comparison_basis":null,"source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC13420647/"],"source_refs":["AIBD-0273-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not measured; outputs did not guide care.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Independent management by clinicians over four months; no control group.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"The system classified one patient as Warm/Wet and recommended maintaining diuretic therapy with intensified monitoring, and classified the other as Cold/Dry and recommended diuretic de-escalation; neither patient needed hospitalization during follow-up.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"None; patients were managed independently of the system.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"The authors ran GPT-5 mini with disease-specific prompts on nursing-facility documentation and set its scores and recommendations beside the care that clinicians gave independently.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Heart Failure Likelihood Score, exacerbation risk and Forrester classification, with narrative management recommendations.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two heart-failure patients in a skilled nursing facility, chosen to show contrasting Forrester hemodynamic quadrants.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two purposively selected cases with no accuracy metric and no comparison group; the senior author founded the platform; the authors call for prospective validation.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0273-reported-measure","evaluation_data":null},{"case_id":"AIBD-0280","case_url":"records/AIBD-0280.html","measure":"successful separation in four staged procedures","reported_result":"successful separation in four staged procedures","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.mediaoffice.abudhabi/en/health/purehealth-performs-gccs-first-separation-of-craniopagus-twins-procedure-in-abu-dhabi/","https://wqow.marketminute.com/article/bizwire-2026-6-19-xrlabs-achieves-world-first-in-ai-designed-implants-for-conjoined-twin-separation"],"source_refs":["AIBD-0280-source-0001","AIBD-0280-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"successful separation in four staged procedures","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"successful separation in four staged procedures","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0280-reported-measure","evaluation_data":null},{"case_id":"AIBD-0492","case_url":"records/AIBD-0492.html","measure":null,"reported_result":null,"baseline":"Internal financing that limited the number of beneficiaries, caused service delays and strained staff","result":"Bank-financed loans with automated eligibility evaluation; MBRHE reports a 30% cut in internal workload for the whole program","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-665","https://www.dib.ae/about-us/news/2025/12/07/mbrhe-and-dib-sign-strategic-partnership-agreement-to-support-housing-finance-services","https://www.emiratesislamic.ae/en/personal-banking/finance/mohammed-bin-rashid-housing-establishment"],"source_refs":["AIBD-0492-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0492-reported-measure","evaluation_data":null},{"case_id":"AIBD-0546","case_url":"records/AIBD-0546.html","measure":"Total falls, unwitnessed bedroom falls, ambulance callouts, hospital transfers, hospital days, safeguarding notices.","reported_result":"total falls fell by 49.2%","baseline":"Pre-installation baseline counts of falls, ambulance callouts and hospital transfers (figures not published)","result":"Total falls down 49.2%, unwitnessed bedroom falls down 58.2%, ambulance callouts down 63.7%, hospital transfers down 79.3%","baseline_value":null,"result_value":null,"unit":null,"effect":"Falls down 49.2%, unwitnessed bedroom falls down 58.2%, ambulance callouts down 63.7%, hospital transfers down 79.3%, hospital days down 72.6%, safeguarding notices down 51.9%.","design":"before/after.","sample":"Six care homes in the Dorset Council and BCP Council areas, 184 planned sensors.","uncertainty":"The deployer ran the comparison and published no baseline counts, periods or resident numbers. There is no control group. One of six homes recorded more falls. The NHS report on the pilot had not been published.","comparison_basis":"measured","source_urls":["https://www.allycares.com/case-study-dorset-icb-night-time-safety-acoustic-monitoring/","https://www.local.gov.uk/case-studies/dorset-council-acoustic-monitoring-care-homes","https://www.bbc.co.uk/news/articles/cyv1qg87yrmo","https://web.archive.org/web/2026/https://www.dorsetcouncil.gov.uk/news/acoustic-monitoring-in-care-homes-pilot-shows-reduction-in-number-of-resident-falls"],"source_refs":["AIBD-0546-source-0004"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The same homes before installation.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Falls down 49.2%, unwitnessed bedroom falls down 58.2%, ambulance callouts down 63.7%, hospital transfers down 79.3%, hospital days down 72.6%, safeguarding notices down 51.9%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Night-time acoustic and infrared monitoring with alerts to staff devices.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Care home records of falls, ambulance callouts, hospital transfers, hospital days and safeguarding notices, compared before and after installation and normalized for different reporting periods.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Total falls, unwitnessed bedroom falls, ambulance callouts, hospital transfers, hospital days, safeguarding notices.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Six care homes in the Dorset Council and BCP Council areas, 184 planned sensors.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The deployer ran the comparison and published no baseline counts, periods or resident numbers. There is no control group. One of six homes recorded more falls. The NHS report on the pilot had not been published.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0546-reported-measure","evaluation_data":null},{"case_id":"AIBD-0547","case_url":"records/AIBD-0547.html","measure":"Time to summarize each submission; quality of the summaries.","reported_result":"About 18 hours of officer summary work against 16 minutes for PlanAI; no discernible difference in summary quality.","baseline":"About 18 hours of officer time to summarize the 320 submissions","result":"16 minutes of PlanAI run time","baseline_value":null,"result_value":null,"unit":null,"effect":"About 18 hours of officer summary work against 16 minutes for PlanAI; no discernible difference in summary quality.","design":"controlled observational.","sample":"Three live public consultations in Greater Cambridge, 320 submissions, 2025.","uncertainty":"The developers ran and reported the test under a paid contract. One officer served as the comparison. The quality assessors and criteria are not described, and no peer-reviewed report exists.","comparison_basis":"measured","source_urls":["https://theconversation.com/the-government-wants-local-authorities-to-embrace-ai-heres-one-way-it-could-work-in-practice-258449","https://www.local.gov.uk/case-studies/greater-cambridge-shared-planning-ai-driven-summarisation-local-plan-consultations","https://web.archive.org/web/2026/https://www.ukauthority.com/articles/ai-planning-tool-expands-its-testing-landscape/"],"source_refs":["AIBD-0547-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"The summary step fell from about 18 hours to 16 minutes. The officer's other work on the 320 submissions, about 42 hours, is not reported as reduced.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"A planning officer processing the same submissions.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"About 18 hours of officer summary work against 16 minutes for PlanAI; no discernible difference in summary quality.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"320 submissions across three consultations.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Officer hours spent on the same submissions were compared with PlanAI run time; summary quality was compared in a later qualitative assessment.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time to summarize each submission; quality of the summaries.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Three live public consultations in Greater Cambridge, 320 submissions, 2025.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The developers ran and reported the test under a paid contract. One officer served as the comparison. The quality assessors and criteria are not described, and no peer-reviewed report exists.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0547-reported-measure","evaluation_data":null},{"case_id":"AIBD-0279","case_url":"records/AIBD-0279.html","measure":"Character error rate and word error rate.","reported_result":"200 million pages deciphered","baseline":"Transkribus PyLaia base model: character error rate 10.3%, word error rate 27.0%","result":"Transkribus Titan Super Model: character error rate 8.0%, word error rate 19.7%; the best LLMs reached 5.7 to 7%","baseline_value":null,"result_value":null,"unit":null,"effect":"Transkribus Titan: CER 8.0%, WER 19.7% (PyLaia: CER 10.3%, WER 27.0%); LLMs: CER 5.7 to 7%.","design":"other.","sample":"50 pages of 18th and 19th century English-language handwritten documents, 500 pages per model per test.","uncertainty":"One English-language corpus; the authors built a competing LLM-based tool; results apply to Transkribus base models without fine-tuning; FamilySearch's own engine was not tested.","comparison_basis":"measured","source_urls":["https://arxiv.org/html/2411.03340v1","https://newsroom.churchofjesuschrist.org/article/familysearch-2024-genealogy-highlights","https://www.transkribus.org/","https://www.geneamusings.com/2026/01/what-to-expect-from-familysearch-in.html","https://www.geneamusings.com/2026/10/amanuensis-monday-1761-deed-of-benjamin.html"],"source_refs":["AIBD-0279-source-0005"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Transkribus PyLaia base model, Transkribus Titan Super Model, and multimodal LLMs from OpenAI, Anthropic and Google.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Transkribus Titan: CER 8.0%, WER 19.7% (PyLaia: CER 10.3%, WER 27.0%); LLMs: CER 5.7 to 7%.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Character and word error rates of machine transcriptions against manually corrected ground truth, each model run ten times on the test set.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Character error rate and word error rate.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"50 pages of 18th and 19th century English-language handwritten documents, 500 pages per model per test.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One English-language corpus; the authors built a competing LLM-based tool; results apply to Transkribus base models without fine-tuning; FamilySearch's own engine was not tested.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0279-reported-measure","evaluation_data":null},{"case_id":"AIBD-0569","case_url":"records/AIBD-0569.html","measure":"Staff accounts of administrative burden and of time spent with people; assessors' review of records made with the tools.","reported_result":"Staff told assessors that AI note-taking tools helped them focus on people during assessments and reduced administrative burden.","baseline":"Staff wrote meeting minutes by hand.","result":"The council reports that automation halved the time spent on minutes and saved an estimated 107 days of staff work.","baseline_value":null,"result_value":null,"unit":null,"effect":"Staff told assessors that AI note-taking tools helped them focus on people during assessments and reduced administrative burden.","design":"qualitative evaluation.","sample":"Wigan Council adult social care, England; assessment published 18 March 2026.","uncertainty":"The report gives no time or cost figures and does not name Agilisys or QuickAction. The AI tools were one part of a service that the regulator rated outstanding for many other reasons.","comparison_basis":"assumed","source_urls":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/wigan-council-quickaction-adult-social-care","https://www.local.gov.uk/case-studies/wigan-council-quickaction-adult-social-care","https://www.cqc.org.uk/care-services/local-authority-assessment-reports/wigan-council/wigan-0326/theme4/learning","https://www.cqc.org.uk/care-services/local-authority-assessment-reports/wigan-council/wigan-0326/theme1/needs","https://www.cqc.org.uk/care-services/local-authority-assessment-reports/wigan-council/wigan-0326/theme4/governance","https://www.agilisys.co.uk/wigan-adult-social-care"],"source_refs":["AIBD-0569-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"No formal comparison; staff compared their work before and after the tools.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"qualitative evaluation.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Staff told assessors that AI note-taking tools helped them focus on people during assessments and reduced administrative burden.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"qualitative evaluation.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Regulatory assessment of the council's adult social care, with interviews of staff, leaders, partners and people who use services, and review of case records.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Staff accounts of administrative burden and of time spent with people; assessors' review of records made with the tools.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Wigan Council adult social care, England; assessment published 18 March 2026.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The report gives no time or cost figures and does not name Agilisys or QuickAction. The AI tools were one part of a service that the regulator rated outstanding for many other reasons.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0569-reported-measure","evaluation_data":null},{"case_id":"AIBD-0607","case_url":"records/AIBD-0607.html","measure":"Change in time-in-note versus control.","reported_result":"cut their time-in-note by 9.5% against the control group","baseline":"Usual-care change in time-in-note","result":"9.5% greater decrease in time-in-note with Nabla","baseline_value":null,"result_value":null,"unit":null,"effect":"Nabla -9.5% (95% CI -17.2% to -1.8%); DAX -1.7% (95% CI -9.4% to +5.9%).","design":"randomized.","sample":"238 outpatient physicians in 14 specialties at UCLA Health; 23,653 Nabla-arm visits.","uncertainty":"Single system, two months, scribe used in under a third of visits, survey outcomes secondary, occasional inaccuracies.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12768499/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13154711/"],"source_refs":["AIBD-0607-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"9.5% less time-in-note for Nabla users.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Usual care (no scribe).","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Nabla -9.5% (95% CI -17.2% to -1.8%); DAX -1.7% (95% CI -9.4% to +5.9%).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Nabla used in 29.5% of visits in its arm.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Epic Signal time-in-note (log-transformed change from baseline); surveys for burnout, task load and accuracy.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Change in time-in-note versus control.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"ClinicalTrials.gov NCT06792890; supplementary appendix.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"238 outpatient physicians in 14 specialties at UCLA Health; 23,653 Nabla-arm visits.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Single system, two months, scribe used in under a third of visits, survey outcomes secondary, occasional inaccuracies.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0607-reported-measure","evaluation_data":null},{"case_id":"AIBD-0350","case_url":"records/AIBD-0350.html","measure":null,"reported_result":null,"baseline":null,"result":"a Speed Range Event alert, then a Fishing Event alert 40 minutes later","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.skylight.global/news/argentina-success-story","https://agendamalvinas.com.ar/noticia/prefectura-detecto-un-buque-asiatico-pescando-ilegalmente-en-aguas-argentinas","https://developer.nvidia.com/blog/ai-model-offers-conservationists-new-tools-to-protect-fisheries-wildlife-at-scale/","https://www.skylight.global/news/impact-report-2025-2026"],"source_refs":["AIBD-0350-source-0002","AIBD-0350-source-0003","AIBD-0350-source-0001","AIBD-0350-source-0004","AIBD-0350-source-0005","AIBD-0350-source-0006","AIBD-0350-source-0007","AIBD-0350-source-0008"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: PNA's capability before Skylight's real-time fishing detection","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Result: a Speed Range Event alert, then a Fishing Event alert 40 minutes later.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0350-reported-measure","evaluation_data":null},{"case_id":"AIBD-0132","case_url":"records/AIBD-0132.html","measure":"Kingston Crown Court jailed him for two years","reported_result":"Kingston Crown Court jailed him for two years","baseline":"it’s possible if he hadn’t been identified using this technology, he could have gone on to abuse this child","result":"his face was identified by Live Facial Recognition cameras attached to the back of a van, alerting officers to his presence","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://southwarknews.co.uk/area/camberwell/sex-offender-caught-in-denmark-hill-thanks-to-facial-recognition-cameras/","https://oecd.ai/en/incidents/2025-05-20-368b","https://www.biometricupdate.com/202507/london-police-mark-1000-arrests-through-live-facial-recognition","https://www.judiciary.uk/wp-content/uploads/2026/04/AC-2024-LON-001764-R-Thompson-and-Carlo-version-for-hand-down-21-04-2026.pdf"],"source_refs":["AIBD-0132-source-0001","AIBD-0132-source-0002","AIBD-0132-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: not being identified (no live facial recognition check)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: it’s possible if he hadn’t been identified using this technology, he could have gone on to abuse this child. Result: his face was identified by Live Facial Recognition cameras attached to the back of a van, alerting officers to his presence.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Kingston Crown Court jailed him for two years","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0132-reported-measure","evaluation_data":null},{"case_id":"AIBD-0081","case_url":"records/AIBD-0081.html","measure":"a saving of R$3,063,774.84","reported_result":"a saving of R$3,063,774.84","baseline":"valor estimado em R$ 15.450.714,80","result":"R$ 12.386.939,96","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/20250114172015id_/https://www.gov.br/cgu/pt-br/assuntos/noticias/2024/12/atuacao-do-robo-alice-segue-auxiliando-auditores-a-realizar-trabalhos-preventivos-que-ajudam-a-evitar-desperdicios-e-irregularidades/","https://convergenciadigital.com.br/governo/robo-da-cgu-rastreia-r-25-bilhoes-em-licitacoes/","https://www.licitacao.net/noticias/alice:-robo-da-cgu-previne-erros-e-desperdicios-de-licitacoes-publicas","https://www.novaleilicitacao.com.br/2025/01/21/atuacao-do-robo-alice-auxilia-cgu-a-rastrear-r-25-bilhoes-em-licitacoes/","https://revista.cgu.gov.br/Revista_da_CGU/article/view/530","https://pmc.ncbi.nlm.nih.gov/articles/PMC10039446/","https://pncp.gov.br/app/contratos/00394544000185/2025/4265"],"source_refs":["AIBD-0081-source-0001","AIBD-0081-source-0002","AIBD-0081-source-0003","AIBD-0081-source-0004","AIBD-0081-source-0005","AIBD-0081-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the hospital's original price estimate for tender 90017/2024","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: valor estimado em R$ 15.450.714,80. Result: R$ 12.386.939,96.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a saving of R$3,063,774.84","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0081-reported-measure","evaluation_data":null},{"case_id":"AIBD-0655","case_url":"records/AIBD-0655.html","measure":"Sensitivity and specificity for fractures.","reported_result":"from 61% to 74%","baseline":"61% sensitivity","result":"74% sensitivity","baseline_value":null,"result_value":null,"unit":null,"effect":"Sensitivity 61% to 74% (BoneView), 74% to 83% (RBfracture), 79% to 94% (Rayvolve); no loss of specificity.","design":"controlled observational.","sample":"16 studies, mostly on BoneView (8) and RBfracture (5).","uncertainty":"Mostly retrospective case-control designs; no UK urgent care studies; readers saw X-rays without patient history. Unassisted readers were less accurate than expected.","comparison_basis":"measured","source_urls":["https://www.nice.org.uk/guidance/htg739/chapter/3-Committee-discussion","https://www.nice.org.uk/guidance/htg739/chapter/1-Recommendations","https://pubmed.ncbi.nlm.nih.gov/33944629/"],"source_refs":["AIBD-0655-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Higher sensitivity by 9 to 15 percentage points.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Unassisted reading by the same clinicians.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sensitivity 61% to 74% (BoneView), 74% to 83% (RBfracture), 79% to 94% (Rayvolve); no loss of specificity.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Reading sessions with and without the software.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sensitivity and specificity of clinicians reading with and without the software against a reference standard.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Sensitivity and specificity for fractures.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"External assessment group report on the NICE website.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"16 studies, mostly on BoneView (8) and RBfracture (5).","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mostly retrospective case-control designs; no UK urgent care studies; readers saw X-rays without patient history. Unassisted readers were less accurate than expected.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0655-reported-measure","evaluation_data":null},{"case_id":"AIBD-0278","case_url":"records/AIBD-0278.html","measure":"estimated its bulk modulus at 169 GPa against a 200 GPa target","reported_result":"estimated its bulk modulus at 169 GPa against a 200 GPa target","baseline":"not stated as a figure in the text (previous generative models)","result":"more than twice as likely to be SUN (stable, unique and new); the synthesized TaCr2O6 measured at 169 GPa against a 200 GPa design target","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11922738/","https://www.microsoft.com/en-us/research/blog/mattergen-a-new-paradigm-of-materials-design-with-generative-ai/","https://europepmc.org/article/MED/42007796","https://pmc.ncbi.nlm.nih.gov/articles/PMC12822528/"],"source_refs":["AIBD-0278-source-0001","AIBD-0278-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"previous generative models for crystal structures (DiffCSP, CDVAE and others) on the same benchmark","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: not stated as a figure in the text (previous generative models). Result: more than twice as likely to be SUN (stable, unique and new); the synthesized TaCr2O6 measured at 169 GPa against a 200 GPa design target.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Estimated bulk modulus of 169 GPa; target 200 GPa","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0278-reported-measure","evaluation_data":null},{"case_id":"AIBD-0572","case_url":"records/AIBD-0572.html","measure":null,"reported_result":null,"baseline":"expenditure on temporary housing exceeding £20 million in the past year","result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/newham-council-establishing-uk-centre-artificial-intelligence-public-sector","https://www.local.gov.uk/case-studies/newham-council-establishing-uk-centre-artificial-intelligence-public-sector","https://loti.london/blog/ai-in-local-government-digigov-expo/"],"source_refs":["AIBD-0572-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0572-reported-measure","evaluation_data":null},{"case_id":"AIBD-0362","case_url":"records/AIBD-0362.html","measure":"Probability of detection and false alarm rate for each class.","reported_result":"detection rates of 95% for shark fins, 96% for seahorses, and 86% for sea cucumbers","baseline":null,"result":"PD shark fin: 95%, seahorse: 96% and sea cucumber: 86%","baseline_value":null,"result_value":null,"unit":null,"effect":"Detection of 95% for shark fins, 96% for seahorses and 86% for sea cucumbers, with class false alarm rates of 2%, 9% and 1%.","design":"other.","sample":"512 test bags scanned on a Rapiscan RTT110 in Homebush, New South Wales: 108 with shark fins, 131 with seahorses, 75 with sea cucumbers and 198 without marine items; most positive bags were synthetic composites.","uncertainty":"No confidence intervals are reported. The test set was sampled from the same population as the training data, and most positive test bags were synthetic composites. Only dead, mostly dried specimens in small luggage were tested. Five authors are employed by Rapiscan, which funded the study and supplied the scanner and staff for analysis and writing.","comparison_basis":"measured","source_urls":["https://www.frontiersin.org/articles/10.3389/focsu.2026.1776978/full","https://www.frontiersin.org/journals/ocean-sustainability/articles/10.3389/focsu.2026.1776978/full","https://news.mongabay.com/2026/06/thinking-how-traffickers-think-study-uses-ai-to-detect-marine-wildlife-smuggling/"],"source_refs":["AIBD-0362-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Known contents of each scanned bag; no comparison with human screeners or detector dogs.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Detection of 95% for shark fins, 96% for seahorses and 86% for sea cucumbers, with class false alarm rates of 2%, 9% and 1%.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Laboratory benchmark: a 3D convolutional network trained on CT scans and synthetic composites, scored on a separate test set sampled from the same population as the training data.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Probability of detection and false alarm rate for each class.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The datasets are not openly available. Requests go to the corresponding author and need approval by Rapiscan and the Australian Department of Climate Change, Energy, the Environment and Water.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"512 test bags scanned on a Rapiscan RTT110 in Homebush, New South Wales: 108 with shark fins, 131 with seahorses, 75 with sea cucumbers and 198 without marine items; most positive bags were synthetic composites.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"No confidence intervals are reported. The test set was sampled from the same population as the training data, and most positive test bags were synthetic composites. Only dead, mostly dried specimens in small luggage were tested. Five authors are employed by Rapiscan, which funded the study and supplied the scanner and staff for analysis and writing.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0362-reported-measure","evaluation_data":null},{"case_id":"AIBD-0543","case_url":"records/AIBD-0543.html","measure":null,"reported_result":null,"baseline":"About 10 minutes of human effort per document","result":"About 90 seconds of automatic redaction, with a 95% success rate","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/birmingham-city-council-foundry","https://www.local.gov.uk/case-studies/birmingham-city-council-foundry","https://assets.publishing.service.gov.uk/media/67eaa234ba01abac8e9fe978/Birmingham_City_Council_Commissioners_second_report.pdf","https://assets.publishing.service.gov.uk/media/692d54af345e31ab14ecf7a2/Birmingham_City_Council_Commissioners_third_report.pdf"],"source_refs":["AIBD-0543-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0543-reported-measure","evaluation_data":null},{"case_id":"AIBD-0564","case_url":"records/AIBD-0564.html","measure":null,"reported_result":null,"baseline":"About £120 per A4 page, upwards of £600 for a complex 5-10 page document, and weeks of work","result":"7-10 pence per document (1 penny per page in another passage of the same account), made in under 30 minutes","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/swindon-borough-council-simply-readable-ai-easy-read-solution","https://www.local.gov.uk/case-studies/swindon-borough-council-simply-readable-ai-easy-read-solution","https://aws.amazon.com/blogs/publicsector/swindon-borough-council-makes-vital-public-information-more-accessible-using-amazon-bedrock/","https://web.archive.org/web/2026/https://www.ukauthority.com/articles/swindon-develops-genai-for-easy-read-documents/"],"source_refs":["AIBD-0564-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0564-reported-measure","evaluation_data":null},{"case_id":"AIBD-0577","case_url":"records/AIBD-0577.html","measure":"Classification accuracy; share of outputs with irrelevant content; processing time.","reported_result":"63% accuracy, against 95% for a simpler method","baseline":"95 per cent","result":"63 per cent","baseline_value":null,"result_value":null,"unit":null,"effect":"63 percent accuracy for the LLM against 95 percent for the simpler method; irrelevant output in 3.8 percent of cases; up to 1.5 hours per output.","design":"other.","sample":"Warwickshire County Council business properties and types; sample size not stated.","uncertainty":"The council evaluated its own proof of concept, and the case study gives no sample size or labeling method. Model labels differ within the text.","comparison_basis":"measured","source_urls":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/warwickshire-county-council-access-food-exploring-llms-citizen-needs","https://www.local.gov.uk/case-studies/warwickshire-county-council-access-food-exploring-llms-citizen-needs"],"source_refs":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/warwickshire-county-council-access-food-exploring-llms-citizen-needs"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Negative: the LLM was less accurate than the method without AI.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"String concatenation without AI.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"63 percent accuracy for the LLM against 95 percent for the simpler method; irrelevant output in 3.8 percent of cases; up to 1.5 hours per output.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Prompt engineering only, with no fine-tuning, on a quantized model run on a laptop.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Accuracy of business classifications from a prompted LLM compared with a rule-based string-concatenation method on council business-property data.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Classification accuracy; share of outputs with irrelevant content; processing time.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Warwickshire County Council business properties and types; sample size not stated.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The council evaluated its own proof of concept, and the case study gives no sample size or labeling method. Model labels differ within the text.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0577-reported-measure","evaluation_data":null},{"case_id":"AIBD-0381","case_url":"records/AIBD-0381.html","measure":null,"reported_result":null,"baseline":"Duran judged that a lawyer would be too costly","result":"He filed an objection drafted with AI and the court annulled the fine","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.webtekno.com/chatgpt-dilekce-trafik-cezasi-h155464.html","https://www.aksam.com.tr/guncel/yapay-zekaya-yazdirdigi-dilekce-ile-trafik-cezasini-iptal-ettirdi/haber-1544041","https://oecd.ai/en/incidents/2025-02-06-4cc2"],"source_refs":["AIBD-0381-source-0002","AIBD-0381-source-0001","AIBD-0381-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0381-reported-measure","evaluation_data":null},{"case_id":"AIBD-0398","case_url":"records/AIBD-0398.html","measure":null,"reported_result":null,"baseline":"20 to 40 minutes per patient for initial review","result":"10 to 15 minutes per patient with AI-assisted image analysis and report editing","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.news.cn/tech/20260831/c1e81084ee504448a9a1db9d8aeee7ad/c.html","https://yicaiglobal.com/news/chinas-ruijin-hospital-huawei-launch-ai-pathology-model"],"source_refs":["AIBD-0398-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0398-reported-measure","evaluation_data":null},{"case_id":"AIBD-0106","case_url":"records/AIBD-0106.html","measure":null,"reported_result":null,"baseline":"one described species, Limnognathia maerski","result":"two species, with the Crozet population described as Limnognathia desmeti","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11836703/"],"source_refs":["AIBD-0106-source-0002","AIBD-0106-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0106-reported-measure","evaluation_data":null},{"case_id":"AIBD-0421","case_url":"records/AIBD-0421.html","measure":"Solar forecast accuracy at a two-hour horizon.","reported_result":"a 5% accuracy gain at a two-hour horizon","baseline":"ECMWF's 2.5% improvement when integrated into PVNet","result":"a 5% improvement in solar forecast accuracy at two hours ahead","baseline_value":null,"result_value":null,"unit":null,"effect":"5% improvement in solar forecast accuracy at two hours ahead.","design":"other.","sample":"UK solar forecasts using EUMETSAT satellite imagery in 11 spectral channels; period not stated.","uncertainty":"The developer gives no test period, error metric or uncertainty range, and the result has not been repeated by an independent party.","comparison_basis":"measured","source_urls":["https://www.openclimatefix.org/insights/meet-cloudcasting-our-newest-forecasting-innovation","https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-1-PDF-E.pdf"],"source_refs":["AIBD-0421-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"5% from the Cloudcasting input, against 2.5% from ECMWF input.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"PVNet with ECMWF forecasts added (2.5% improvement).","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"5% improvement in solar forecast accuracy at two hours ahead.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Developer's comparison of PVNet solar forecasts with and without Cloudcasting predicted satellite images, and with ECMWF cloud forecasts as an alternative input.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Solar forecast accuracy at a two-hour horizon.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Open-source code at https://github.com/openclimatefix/sat_pred, https://github.com/openclimatefix/cloudcasting-app and https://github.com/alan-turing-institute/cloudcasting; EUMETSAT SEVIRI data on Google Cloud public datasets.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"UK solar forecasts using EUMETSAT satellite imagery in 11 spectral channels; period not stated.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The developer gives no test period, error metric or uncertainty range, and the result has not been repeated by an independent party.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0421-reported-measure","evaluation_data":null},{"case_id":"AIBD-0143","case_url":"records/AIBD-0143.html","measure":"Study 1: precipitation RMSE and categorical scores at lead times of 1 to 5 days. Study 2: 2 m temperature mean absolute error and precipitation category skill.","reported_result":"5–15% lower errors than IFS","baseline":null,"result":"errors 5 to 15% lower in the medium range (out to 15 days)","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: AIFS RMSE at 120 hours was roughly equal to IFS RMSE at 24 hours; AIFS scored better for heavy rain above 25 mm but over-forecast light rain and scored worse on SEEPS. Study 2: AIFS temperature mean absolute error about 15.7% lower than the IFS control forecast; AIFS better for moderate rain and worse for heavy rain.","design":"Study 1: controlled observational. Study 2: controlled observational.","sample":"Study 1: 2,415 stations in China, April 2024 to March 2025. Study 2: 117 stations in Inner Mongolia, June to August 2026.","uncertainty":"Study 1: The period falls mostly before AIFS Single v1 became operational, and the study covers one country. The authors link the weak AIFS extremes to weak vertical velocity and water vapor transport. One sentence of the abstract states the RMSE comparison the other way round from the results section. Study 2: One region and one season; the authors compare operational products and state that their findings apply only to the versions, region and season evaluated.","comparison_basis":"measured","source_urls":["https://www.ecmwf.int/en/about/media-centre/news/2026/2025-forecast-evaluation","https://www.ecmwf.int/node/29308","https://www.ecmwf.int/en/newsletter/187/news/forecast-performance-2025","https://www.weather.gov/news/261102-AI-Hurricane-Forecasting","https://confluence.ecmwf.int/display/FCST/202501+-+Windstorm+-+Eowyn","https://arxiv.org/abs/2608.09959","https://journals.ametsoc.org/view/journals/wefo/40/9/WAF-D-24-0227.1.xml","https://doi.org/10.3390/atmos17111053"],"source_refs":["AIBD-0143-source-0007","AIBD-0143-source-0008"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: lower precipitation RMSE at 1 to 5 days, with weaker extremes. Study 2: about 15.7% lower temperature error than the IFS control forecast.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: ECMWF's physics-based IFS over the same period and stations. Study 2: the IFS control forecast and the WeatherNext 2 ensemble mean.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: controlled observational. Study 2: controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: AIFS RMSE at 120 hours was roughly equal to IFS RMSE at 24 hours; AIFS scored better for heavy rain above 25 mm but over-forecast light rain and scored worse on SEEPS. Study 2: AIFS temperature mean absolute error about 15.7% lower than the IFS control forecast; AIFS better for moderate rain and worse for heavy rain.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: daily AIFS and IFS forecasts over twelve months. Study 2: operational AIFS forecasts over one summer flood season.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: controlled observational. Study 2: controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Verification of AIFS and IFS precipitation forecasts against station observations and ERA5, with RMSE, mean error, threat scores, equitable threat scores and SEEPS. Study 2: Verification of 2 m temperature and 6-hour precipitation at 6 to 84 hours with Murphy error decomposition and block-bootstrap tests.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: precipitation RMSE and categorical scores at lead times of 1 to 5 days. Study 2: 2 m temperature mean absolute error and precipitation category skill.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 2,415 stations in China, April 2024 to March 2025. Study 2: 117 stations in Inner Mongolia, June to August 2026.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: The period falls mostly before AIFS Single v1 became operational, and the study covers one country. The authors link the weak AIFS extremes to weak vertical velocity and water vapor transport. One sentence of the abstract states the RMSE comparison the other way round from the results section. Study 2: One region and one season; the authors compare operational products and state that their findings apply only to the versions, region and season evaluated.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0143-reported-measure","evaluation_data":null},{"case_id":"AIBD-0151","case_url":"records/AIBD-0151.html","measure":"found seven sperm, two of them motile, in 2.5 million images of a “sperm-free” sample","reported_result":"found seven sperm, two of them motile, in 2.5 million images of a “sperm-free” sample","baseline":"several manual sperm searches, and two surgical procedures to extract sperm","result":"In about two hours, STAR scanned 2.5 million images, identifying 2 viable sperm cells","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.eurekalert.org/news-releases/1104028","https://www.columbiadoctors.org/specialties/obstetrics-gynecology/our-services/columbia-university-fertility-center/our-services/sperm-tracking-and-recovery-star","https://doi.org/10.1016/S0140-6736(25)01623-X","https://www.washingtonpost.com/health/2025/06/27/ai-enabled-search-viable-sperm-infertility/","https://time.com/7291154/doctors-report-first-pregnancy-new-ai-procedure/","https://www.cnn.com/2025/07/03/health/ai-male-infertility-sperm-wellness","https://www.femtechworld.co.uk/insight/doctors-report-first-pregnancy-using-ai-system-to-detect-sperm-in-men-previously-considered-infertile-aift25/","https://time.com/collections/best-inventions-2025/7318439/columbia-university-star-method/","https://www.columbiaspectator.com/news/2025/10/26/this-ai-imaging-tech-from-columbia-researchers-is-solving-male-infertility/","https://economictimes.indiatimes.com/news/india/how-ai-helped-an-infertile-couple-conceive-after-19-years-with-just-2-sperm/articleshow/125051307.cms","https://www.freepressjournal.in/tech/how-ai-helped-this-couple-get-pregnant-after-19-year-wait","https://www.republicworld.com/tech/after-19-years-of-infertility-ai-finds-2-sperm-from-25-million-images-couple-welcomes-miracle-baby","https://www.thedailyjagran.com/technology/ai-helps-couple-conceive-after-20-years-of-infertility-a-medical-breakthrough-from-columbia-university-10277754","https://in.mashable.com/tech/102002/after-19-years-of-failed-ivf-attempts-ai-helps-couple-conceive-in-groundbreaking-first","https://studyfinds.com/ai-detects-invisible-sperm/","https://biz.chosun.com/en/en-science/2025/11/01/DUEJDMOPVVAINJVKBQ2ERLB6DU/","https://www.dongascience.com/en/news/74880","https://en.ilsole24ore.com/art/first-baby-thanks-to-artificial-intelligence-and-robot-breakthrough-in-male-infertility-AH8Kc0WD","https://interestingengineering.com/science/breakthrough-ai-helps-infertile-man","https://medicaldialogues.in/obstetrics-gynaecology/news/first-pregnancy-with-ai-guided-sperm-recovery-method-developed-at-columbia-study-160260","https://www.bbc.com/future/article/20260429-finding-hidden-sperm-new-technology-offers-hope-to-men-previously-told-they-were-infertile","https://nypost.com/2026/05/13/health/inside-the-new-ivf-tech-giving-rise-to-ai-babies/","https://www.dongascience.com/en/news/77712","https://www.thestar.com.my/tech/tech-news/2026/08/13/ai-is-finding-sperm-where-doctors-couldnt","https://www.medicaldaily.com/ai-sperm-detection-azoospermia-star-method-evidence-477211","https://inciid.org/star-sperm-tracking-and-recovery-system/","https://www.columbiadoctors.org/news/columbia-fertility-looks-stars-help-men-infertility","https://www.vagelos.columbia.edu/about-us/columbia-medicine-magazine/archives/fall-2025/clinical-advances/columbia-develops-star-technology-men-infertility","https://www.nyp.org/advances/podcast/building-an-ai-powered-system-to-improve-fertility-success","https://patents.google.com/patent/US20250264398A1/en","https://patents.google.com/patent/WO2026019779A1/en","https://doi.org/10.1097/MOU.0000000000001430","https://doi.org/10.3389/fendo.2026.1910448"],"source_refs":["AIBD-0151-source-0002","AIBD-0151-source-0003","AIBD-0151-source-0004","AIBD-0151-source-0005","AIBD-0151-source-0006","AIBD-0151-source-0007","AIBD-0151-source-0008","AIBD-0151-source-0009","AIBD-0151-source-0010","AIBD-0151-source-0011","AIBD-0151-source-0012","AIBD-0151-source-0013","AIBD-0151-source-0014","AIBD-0151-source-0015","AIBD-0151-source-0016","AIBD-0151-source-0017","AIBD-0151-source-0018","AIBD-0151-source-0019","AIBD-0151-source-0020","AIBD-0151-source-0021","AIBD-0151-source-0022","AIBD-0151-source-0023","AIBD-0151-source-0024","AIBD-0151-source-0025","AIBD-0151-source-0026","AIBD-0151-source-0027","AIBD-0151-source-0001","AIBD-0151-source-0028","AIBD-0151-source-0029","AIBD-0151-source-0030","AIBD-0151-source-0031","AIBD-0151-source-0032","AIBD-0151-source-0033"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"found 7 sperm (2 motile) in 2.5m images of a “sperm-free” sample","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"found 7 sperm (2 motile) in 2.5m images of a “sperm-free” sample","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0151-reported-measure","evaluation_data":null},{"case_id":"AIBD-0499","case_url":"records/AIBD-0499.html","measure":null,"reported_result":null,"baseline":"Teachers found it hard to give each student one-to-one feedback.","result":"The makers state that AI Helpy gives tailored feedback and reduces teacher workload; no measurement is reported.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://web.archive.org/web/20251110192151/https://elice.io/en/newsroom/eliceschool-aidt","https://journal.kace.re.kr/_common/do.php?a=full&b=12&bidx=4277&aidx=47378","https://www.jksmea.org/journal/view.html?doi=10.63311/mathedu.25.6443"],"source_refs":["AIBD-0499-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0499-reported-measure","evaluation_data":null},{"case_id":"AIBD-0125","case_url":"records/AIBD-0125.html","measure":"a fugitive convict wanted to serve 10 years, 1 month and 10 days for robbery","reported_result":"a fugitive convict wanted to serve 10 years, 1 month and 10 days for robbery","baseline":"An unidentified man without ID","result":"The man identified as a fugitive convict wanted to serve 10 years, 1 month and 10 days","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.ntv.com.tr/turkiye/firari-hukumlu-yuz-tanima-sistemiyle-yakayi-ele-verdi,U5c9NcXxdUWolAOeyn13sA","https://www.haberturk.com/yuz-tanima-sistemiyle-yakayi-eleverdi-3770669","https://www.haberler.com/3-sayfa/yuz-tanima-sistemiyle-yakalanan-hukumlu-kimligim-yok-dedi-18423978-haberi"],"source_refs":["AIBD-0125-source-0002","AIBD-0125-source-0001","AIBD-0125-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a fugitive convict wanted to serve 10 years, 1 month and 10 days for robbery","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a fugitive convict wanted to serve 10 years, 1 month and 10 days for robbery","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0125-reported-measure","evaluation_data":null},{"case_id":"AIBD-0566","case_url":"records/AIBD-0566.html","measure":null,"reported_result":null,"baseline":"A slow process that often produced scans with errors","result":"Scan and image processing 50 per cent faster, with better scan quality; the council does not say how it measured this","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/tunbridge-wells-borough-council-amelia-scott-collecting-tomorrow","https://www.local.gov.uk/case-studies/tunbridge-wells-borough-council-amelia-scott-collecting-tomorrow"],"source_refs":["AIBD-0566-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0566-reported-measure","evaluation_data":null},{"case_id":"AIBD-0092","case_url":"records/AIBD-0092.html","measure":"Probability of detection and false-alarm ratio across lead times and flood thresholds.","reported_result":"some early-action plans could extend from four to seven days without much loss of forecast skill","baseline":"a lead time of at most four days","result":"skill sufficient for early action beyond four days in frequently flooded regions, which suggests a seven-day lead time; Flood Hub's false-alarm ratio is about 0.8 at seven days against about 0.5 for GloFAS","baseline_value":null,"result_value":null,"unit":null,"effect":"Flood Hub and GloFAS keep enough skill beyond a four-day lead time in frequently flooded regions; at seven days Flood Hub's false-alarm ratio is about 0.8 against about 0.5 for GloFAS.","design":"other.","sample":"River stations and districts in Mali; discharge analysis and impact analysis from 1 January 2016 to 31 December 2022.","uncertainty":"Few gauges with short overlap with observations (2016 to 2018); impact records are sparse and inconsistent; no correctly predicted floods in the northern districts evaluated.","comparison_basis":"measured","source_urls":["https://eartharxiv.org/repository/view/11374/","https://doi.org/10.1111/jfr3.70247"],"source_refs":["https://eartharxiv.org/repository/object/11374/download/20631/"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Flood Hub detects more floods than GloFAS at return-period thresholds but with more false alarms; the authors describe different strengths for each model.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"GloFAS v3.0 and v4.0 and the Mali Red Cross protocol's current trigger model.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Flood Hub and GloFAS keep enough skill beyond a four-day lead time in frequently flooded regions; at seven days Flood Hub's false-alarm ratio is about 0.8 against about 0.5 for GloFAS.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective verification of archived reforecasts against river-gauge discharge and district-level flood-impact records, with exceedances merged within a ten-day action window.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Probability of detection and false-alarm ratio across lead times and flood thresholds.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Impact-data code at github.com/rodekruis/disaster-impact-data-analysis; GloFAS and Google reforecasts are public; Mali gauge data are held under a non-disclosure agreement.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"River stations and districts in Mali; discharge analysis and impact analysis from 1 January 2016 to 31 December 2022.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Few gauges with short overlap with observations (2016 to 2018); impact records are sparse and inconsistent; no correctly predicted floods in the northern districts evaluated.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0092-reported-measure","evaluation_data":null},{"case_id":"AIBD-0072","case_url":"records/AIBD-0072.html","measure":"without the eBL platform the edition could not have used 21 manuscripts in 31 fragments","reported_result":"without the eBL platform the edition could not have used 21 manuscripts in 31 fragments","baseline":"impossible for this edition to use 21 manuscripts, consisting of 31 fragments","result":"21 manuscripts, consisting of 31 fragments, most of them previously published only on eBL","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.cambridge.org/core/product/identifier/S0021088924000238/type/journal_article","https://www.lmu.de/en/about-lmu/structure/central-university-administration/media-relations-and-communications/press-room/press-release/lmu-professor-enrique-jimenez-discovered-hymn-to-babylon-8e671888.html","https://www.sciencealert.com/classic-hymn-deciphered-from-ancient-babylonian-library","https://web.archive.org/web/20260926093944/https://archaeology.org/issues/january-february-2026/collection/hymn-to-babylon/top-10-discoveries-of-2025/","https://www.cambridge.org/core/journals/iraq/article/gods-in-the-classroom-religion-and-education-in-1st-millennium-bce-babylonia/EB059EE81804DCBC1848CACBD6298E0B"],"source_refs":["AIBD-0072-source-0001","AIBD-0072-source-0002","AIBD-0072-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: an edition prepared without the eBL platform","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: impossible for this edition to use 21 manuscripts, consisting of 31 fragments. Result: 21 manuscripts, consisting of 31 fragments, most of them previously published only on eBL.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"without the eBL platform the edition could not have used 21 manuscripts in 31 fragments","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0072-reported-measure","evaluation_data":null},{"case_id":"AIBD-0382","case_url":"records/AIBD-0382.html","measure":"86.6 kg of skunk in 119 packages","reported_result":"86.6 kg of skunk in 119 packages","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.sozcu.com.tr/86-kilo-uyusturucuyu-narkotik-kopegi-buldu-p151719","https://yerel-haberler.haberturk.com/sirnak-haberleri/durdurulan-aractaki-86-kilo-uyusturucuyu-hektor-buldu-112096987","https://www.diken.com.tr/sirnakta-86-kilo-uyusturucuyla-yakalanan-uc-avukat-tutuklandi","https://oecd.ai/en/incidents/2025-03-17-8c32","https://www.hurriyetdailynews.com/turkey-using-ai-software-asena-in-fight-against-drugs-173912"],"source_refs":["AIBD-0382-source-0001","AIBD-0382-source-0002","AIBD-0382-source-0003","AIBD-0382-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"86.6 kg of skunk in 119 packages","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"86.6 kg of skunk in 119 packages","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0382-reported-measure","evaluation_data":null},{"case_id":"AIBD-0078","case_url":"records/AIBD-0078.html","measure":"101 minutes against a 199.7-minute average","reported_result":"101 minutes against a 199.7-minute average","baseline":"199.7 minutes on average, over an average distance of 13.6 kilometers","result":"101 minutes over 13.7 miles","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/33422715/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12006056/"],"source_refs":["AIBD-0078-source-0001","AIBD-0078-source-0002","AIBD-0078-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the average interhospital transfer time for similar intracerebral hemorrhage cases, taken from a published New York City hospital-system cohort","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 199.7 minutes. Result: 101 minutes.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"transfer in 101 min vs 199.7 average; mRS 1 at one year","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0078-reported-measure","evaluation_data":null},{"case_id":"AIBD-0126","case_url":"records/AIBD-0126.html","measure":"two arrested within about a day and Rs 79.5 lakh of the Rs 80 lakh recovered","reported_result":"two arrested within about a day and Rs 79.5 lakh of the Rs 80 lakh recovered","baseline":"Rs 80 lakh taken from the courier at gunpoint","result":"Rs 79.5 lakh recovered from Samir's home","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.rediff.com/news/report/israeli-tech-helps-delhi-police-crack-rs-80-lakh-robbery-2-held/20250319.htm","https://timesofindia.indiatimes.com/city/delhi/israeli-face-recognition-software-ends-robbers-run-within-24-hrs/articleshow/119222102.cms","https://indianexpress.com/article/cities/delhi/cctv-footage-face-recognition-tech-how-delhi-cops-nabbed-two-men-who-stole-rs-80-lakh-from-trader-9895565/","https://www.hindustantimes.com/india-news/israeli-facial-recognition-tech-helps-delhi-police-crack-rs-80-lakh-robbery-case-2-held-101742373825793.html","https://www.newindianexpress.com/cities/delhi/2025/Mar/20/israeli-facial-recognition-tech-helps-delhi-police-crack-rs-80-lakh-robbery-case-two-held","https://web.archive.org/web/20220817120915/https://internetfreedom.in/delhi-polices-frt-use-is-80-accurate-and-100-scary/"],"source_refs":["AIBD-0126-source-0002","AIBD-0126-source-0003","AIBD-0126-source-0004","AIBD-0126-source-0001","AIBD-0126-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"two arrested within about a day and Rs 79.5 lakh of the Rs 80 lakh recovered","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"two arrested within about a day and Rs 79.5 lakh of the Rs 80 lakh recovered","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0126-reported-measure","evaluation_data":null},{"case_id":"AIBD-0608","case_url":"records/AIBD-0608.html","measure":"Work exhaustion/interpersonal disengagement score; hours a day on notes.","reported_result":"Time spent on notes fell by 0.36 hours a day","baseline":"Usual documentation","result":"0.36 fewer hours a day on notes","baseline_value":null,"result_value":null,"unit":null,"effect":"-0.44 points exhaustion (95% CI -0.62 to -0.25); -0.36 hours a day on notes (95% CI -0.55 to -0.17).","design":"randomized.","sample":"66 practitioners in UW Health ambulatory clinics in Wisconsin and Illinois; 71,487 notes.","uncertainty":"Small sample, one health system, survey outcomes, fulfillment change not significant, no patient-reported outcomes.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12858090/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13154711/"],"source_refs":["AIBD-0608-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"0.36 fewer hours a day on notes.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Periods without scribe access.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"-0.44 points exhaustion (95% CI -0.62 to -0.25); -0.36 hours a day on notes (95% CI -0.55 to -0.17).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"38% of notes generated with the scribe.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Stanford Professional Fulfillment Index; EHR audit logs for note time; PDSQI-9 note quality; coder review of billing codes.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Work exhaustion/interpersonal disengagement score; hours a day on notes.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"ClinicalTrials.gov NCT06517082; statistical analysis plan and code in a public Git repository.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"66 practitioners in UW Health ambulatory clinics in Wisconsin and Illinois; 71,487 notes.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small sample, one health system, survey outcomes, fulfillment change not significant, no patient-reported outcomes.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0608-reported-measure","evaluation_data":null},{"case_id":"AIBD-0145","case_url":"records/AIBD-0145.html","measure":"£3.4 billion of additional annual yield, about £4.6 billion in 2024–25","reported_result":"£3.4 billion of additional annual yield, about £4.6 billion in 2024–25","baseline":"weeks to find evidence of potential evasion","result":"seconds to search more than a billion pieces of data","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.nao.org.uk/wp-content/uploads/2015/02/Increasing-the-effectiveness-of-tax-collection-a-stocktake-of-progress-since-2010.pdf","https://assets.publishing.service.gov.uk/media/687e5d318adf4250705c96d8/HMRC_annual_report_and_accounts_2024_to_2025.pdf","https://assets.publishing.service.gov.uk/media/5a78b230ed915d07d35b1b53/annual-report-accounts-1011.pdf","https://www.pinsentmasons.com/out-law/news/hmrc-big-data-tax-revenue-boost-4-6-billion","https://assets.publishing.service.gov.uk/media/6a4f66c406e7256c331df8a2/HMRC_annual_report_and_accounts_2025_to_2026.pdf"],"source_refs":["AIBD-0145-source-0001","AIBD-0145-source-0002","AIBD-0145-source-0003","AIBD-0145-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"£3.4 billion of additional annual yield, about £4.6 billion in 2024–25","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"£3.4 billion of additional annual yield, about £4.6 billion in 2024–25","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0145-reported-measure","evaluation_data":null},{"case_id":"AIBD-0046","case_url":"records/AIBD-0046.html","measure":"three small-protein cryo-EM structures at 2.29 to 2.50 angstroms","reported_result":"three small-protein cryo-EM structures at 2.29 to 2.50 angstroms","baseline":"determined the structure of the cargo protein (GFP) at a local resolution of 3.8 Å but failed to provide a reliable atomic interpretation","result":"Aequorea coerulescens GFP (2.29 Å)","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://europepmc.org/article/PMC/PMC13043659","https://www.nature.com/articles/s41586-023-06415-8","https://static-content.springer.com/esm/art%3A10.1038%2Fs41467-026-69941-9/MediaObjects/41467_2026_69941_MOESM3_ESM.pdf","https://www.rcsb.org/structure/9UAQ"],"source_refs":["AIBD-0046-source-0001","AIBD-0046-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"earlier DARPin-containing cage fiducials for small proteins","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: determined the structure of the cargo protein (GFP) at a local resolution of 3.8 Å but failed to provide a reliable atomic interpretation. Result: Aequorea coerulescens GFP (2.29 Å).","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"four small-protein structures at 2.29 to 2.62 angstroms","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0046-reported-measure","evaluation_data":null},{"case_id":"AIBD-0646","case_url":"records/AIBD-0646.html","measure":"Adenoma detection rate.","reported_result":"found adenomas in 52.3% of patients against 36.1% with standard colonoscopy","baseline":"Adenoma detection rate of 36.1%","result":"Adenoma detection rate of 52.3%","baseline_value":null,"result_value":null,"unit":null,"effect":"52.3% vs 36.1%; polyp detection 72.2% vs 54.5%.","design":"randomized.","sample":"998 patients at two tertiary hospitals in Seoul, May 2023 to April 2025.","uncertainty":"Two hospitals; no gain for sessile serrated lesions or cancers.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12800677/"],"source_refs":["AIBD-0646-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"16.2 percentage points in adenoma detection.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Conventional colonoscopy.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"52.3% vs 36.1%; polyp detection 72.2% vs 54.5%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"System used throughout colonoscopy in the intervention group.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Histologically confirmed adenomas per patient.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Adenoma detection rate.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Registered as KCT0009664.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"998 patients at two tertiary hospitals in Seoul, May 2023 to April 2025.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two hospitals; no gain for sessile serrated lesions or cancers.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0646-reported-measure","evaluation_data":null},{"case_id":"AIBD-0128","case_url":"records/AIBD-0128.html","measure":null,"reported_result":null,"baseline":"the forged card carried the impersonator's own photograph, so a check of the card alone would show a match","result":"the app compared her live photo with the photo registered for the hall ticket number and flagged a mismatch","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.thehindu.com/news/cities/bangalore/cet-impersonator-caught-using-facial-recognition-app-probe-ordered/article69461334.ece","https://timesofindia.indiatimes.com/city/bengaluru/cet-facial-recognition-app-nails-fake-candidate-in-bengaluru/articleshow/120388988.cms","https://www.newindianexpress.com/cities/bengaluru/2025/Apr/18/qr-code-facial-recognition-app-exposes-impersonator-in-bengaluru","https://www.etvbharat.com/en/!state/karnataka-examination-authoritys-new-facial-recognition-system-helps-identify-a-proxy-candidate-investigation-underway-enn25041801048","https://timesofindia.indiatimes.com/city/bengaluru/karnataka-common-entrance-test-set-to-begin-with-new-security-norms-for-candidates/articleshow/120288622.cms","https://oecd.ai/en/incidents/2025-04-14-e47b","https://www.etvbharat.com/en/state/karnataka-education-authority-to-conduct-k-cet-on-april-23-and-24-enn26042204512"],"source_refs":["AIBD-0128-source-0001","AIBD-0128-source-0002","AIBD-0128-source-0003","AIBD-0128-source-0004","AIBD-0128-source-0005","AIBD-0128-source-0006","AIBD-0128-source-0007"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0128-reported-measure","evaluation_data":null},{"case_id":"AIBD-0583","case_url":"records/AIBD-0583.html","measure":"Of 1,485,924 images, 88,863 remained for human review. Manual review took 265 hours; the hybrid workflow was estimated at 43.5 hours. Hybrid and manual review found 217 and 378 independent snake events, respectively.","reported_result":"88,863 images remained for human review, a reported 94.0% reduction","baseline":"265 hours of measured manual review; 378 independent snake detections.","result":"43.5 hours estimated for the hybrid workflow, including 27.7 hours of measured compute; 217 independent snake detections.","baseline_value":null,"result_value":null,"unit":null,"effect":"At threshold 0.05, the hybrid workflow left 88,863 images for human review and had an estimated total duration of 43.5 hours. It found 217 independent events compared with 378 manually.","design":"Comparison of manual and AI-assisted review of the same image dataset, using the same reviewer. The paper does not report a randomized or blinded workflow comparison.","sample":"42 cameras at three South Florida sites produced 1,485,924 images between June 20 and July 18, 2022.","uncertainty":"Human review within the hybrid workflow was estimated. The same reviewer assessed both methods. Ground truth combined their findings and could miss snakes overlooked by both. Results depend on species size, camera conditions, hardware, and threshold. Image recall and independent-event detection use different denominators.","comparison_basis":"assumed","source_urls":["https://doi.org/10.1002/rse2.70009","https://github.com/agentmorris/MegaDetector/blob/51270ebad6131b9f2eb7b200ecc706ab616d698c/megadetector.md"],"source_refs":["AIBD-0583-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed"},"evaluations[].design":{"state":"value_present","review_state":"reviewed"},"evaluations[].effect":{"state":"value_present","review_state":"reviewed"},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed"},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed"},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Of 1,485,924 images, 88,863 remained for human review. Manual review took 265 hours; the hybrid workflow was estimated at 43.5 hours. Hybrid and manual review found 217 and 378 independent snake events, respectively.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed"},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed"}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0583-reported-measure","evaluation_data":{"images_total":1485924,"images_for_hybrid_review":88863,"reported_reduction_percent":94.0,"manual_hours_measured":265,"computer_hours_measured":27.7,"hybrid_total_hours_estimated":43.5,"hybrid_human_hours_estimated_approx":16,"independent_events_manual":378,"independent_events_hybrid":217,"independent_events_combined":447,"snake_image_recall_manual_percent":96.2,"snake_image_recall_hybrid_percent":17.1}},{"case_id":"AIBD-0129","case_url":"records/AIBD-0129.html","measure":null,"reported_result":null,"baseline":null,"result":"該案在接獲陳情後24小時內即查到違規的民眾","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://n.yam.com/Article/20250429841522","https://news.ltn.com.tw/news/society/breakingnews/5026699","https://tw.news.yahoo.com/%E5%85%AB%E5%BE%B7%E4%BB%8B%E5%A3%BD%E8%B7%AF%E5%99%AA%E9%9F%B3%E8%BB%8A%E5%BC%95%E6%B0%91%E6%80%A8-%E7%92%B0%E8%AD%A624%E5%B0%8F%E6%99%82%E8%BF%BD%E6%9F%A5%E5%88%B0%E6%A1%88%E8%A3%81%E7%BD%B0-161300728.html","https://n.yam.com/Article/20250429591252","https://www.ctee.com.tw/news/20260108701305-431401"],"source_refs":["AIBD-0129-source-0001","AIBD-0129-source-0002","AIBD-0129-source-0003","AIBD-0129-source-0004","AIBD-0129-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: tracking noise vehicles without the mobile sound cameras (the bureau's own assertion; no baseline figure given)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Result: 該案在接獲陳情後24小時內即查到違規的民眾.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0129-reported-measure","evaluation_data":null},{"case_id":"AIBD-0349","case_url":"records/AIBD-0349.html","measure":"Alerts reached more than 16 million Android users","reported_result":"Alerts reached more than 16 million Android users","baseline":null,"result":"AEA detected the event 5.31 seconds after its initiation, alerting over 16 million users.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.ebi.ac.uk/europepmc/webservices/rest/PMC13558587/fullTextXML","https://europepmc.org/article/PMC/PMC13558587","https://doi.org/10.1038/s41467-026-76413-7","https://europepmc.org/abstract/MED/40674485","https://www.bbc.com/news/articles/c77v2kx304go"],"source_refs":["AIBD-0349-source-0002","AIBD-0349-source-0003","AIBD-0349-source-0004","AIBD-0349-source-0005","AIBD-0349-source-0006","AIBD-0349-source-0007","AIBD-0349-source-0008","AIBD-0349-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"local seismic networks (AFAD stations)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Result: AEA detected the event 5.31 seconds after its initiation, alerting over 16 million users.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"More than 16 million phones received alerts","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0349-reported-measure","evaluation_data":null},{"case_id":"AIBD-0236","case_url":"records/AIBD-0236.html","measure":null,"reported_result":null,"baseline":"Claude Sonnet 3.7 finds it 3 out of 100 runs and Claude Sonnet 3.5 does not find it in 100 runs","result":"o3 finds the kerberos authentication vulnerability in the benchmark in 8 of the 100 runs","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://sean.heelan.io/2025/05/22/how-i-used-o3-to-find-cve-2025-37899-a-remote-zeroday-vulnerability-in-the-linux-kernels-smb-implementation/","https://www.cve.org/CVERecord?id=CVE-2025-37899","https://linuxiac.com/chatgpt-o3-model-found-remote-zeroday-in-linux-kernel-code/","https://github.com/torvalds/linux/commit/2fc9feff45d92a92cd5f96487655d5be23fb7e2b","https://github.com/gregkh/linux/commit/931dc8a3670f71c45c0b1379ea4e92dafbda1aca"],"source_refs":["AIBD-0236-source-0001","AIBD-0236-source-0002","AIBD-0236-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Claude Sonnet 3.7 and Claude Sonnet 3.5 on the same 100-run benchmark (the known ksmbd Kerberos bug)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: Claude Sonnet 3.7 finds it 3 out of 100 runs and Claude Sonnet 3.5 does not find it in 100 runs. Result: o3 finds the kerberos authentication vulnerability in the benchmark in 8 of the 100 runs.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0236-reported-measure","evaluation_data":null},{"case_id":"AIBD-0253","case_url":"records/AIBD-0253.html","measure":"a magnitude 4.4 earthquake, during which trains on the Rome-Naples line were stopped","reported_result":"a magnitude 4.4 earthquake, during which trains on the Rome-Naples line were stopped","baseline":"traditional approaches that shut down entire lines","result":"restrictions applied only to the railway segments where strong shaking is predicted","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://nhess.copernicus.org/articles/26/299/2026/","https://it.euronews.com/viaggi/2025/07/15/treni-che-si-fermano-da-soli-litalia-inaugura-lallerta-sismica-automatica-in-ferrovia","https://www.ansa.it/sito/notizie/economia/2025/07/14/-fs-su-av-roma-napoli-attivo-primo-sistema-allerta-sismica-_eebe792d-b886-4597-af5d-bd9ce79fc8e4.html","https://oecd.ai/en/incidents/2025-07-14-dd98","https://www.wired.it/article/sistema-italiano-ferma-treni-prima-terremoto-roma-napoli/","https://www.quotidiano.net/napoli/terremoto-campi-flegrei-magnitudo-paura-napoli-tgdo9125"],"source_refs":["AIBD-0253-source-0001","AIBD-0253-source-0002","AIBD-0253-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a magnitude 4.4 earthquake, during which trains on the Rome-Naples line were stopped","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a magnitude 4.4 earthquake, during which trains on the Rome-Naples line were stopped","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0253-reported-measure","evaluation_data":null},{"case_id":"AIBD-0603","case_url":"records/AIBD-0603.html","measure":"Odds of fatal or serious injury.","reported_result":"20% lower odds of a fatal or serious injury (adjusted odds ratio 0.80, 95% CI 0.64 to 0.996)","baseline":"Odds of fatal or serious pedestrian injury without AEB","result":"Adjusted odds ratio 0.80 (95% CI 0.64 to 0.996)","baseline_value":null,"result_value":null,"unit":null,"effect":"Pedestrians: adjusted odds ratio 0.80 (0.64 to 0.996); cyclists: 0.91 (0.74 to 1.14), a nonsignificant result.","design":"controlled observational.","sample":"4,131 pedestrians and 6,659 cyclists in police-reported crashes with six best-selling car models, Japan, 2016 to 2019.","uncertainty":"Cross-sectional; counts only crashes that occurred; the upper confidence bound is near 1.","comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/40349424/"],"source_refs":["AIBD-0603-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"20% lower odds of fatal or serious pedestrian injury.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Same models without AEB.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Pedestrians: adjusted odds ratio 0.80 (0.64 to 0.996); cyclists: 0.91 (0.74 to 1.14), a nonsignificant result.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"AEB fitted to the striking car.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Multivariable logistic regression of injury severity (fatal or serious against minor or none).","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Odds of fatal or serious injury.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"National police crash data; code not public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"4,131 pedestrians and 6,659 cyclists in police-reported crashes with six best-selling car models, Japan, 2016 to 2019.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Cross-sectional; counts only crashes that occurred; the upper confidence bound is near 1.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0603-reported-measure","evaluation_data":null},{"case_id":"AIBD-0076","case_url":"records/AIBD-0076.html","measure":"4x4 complex-valued matrices with 48 scalar multiplications, improving on the 49","reported_result":"4x4 complex-valued matrices with 48 scalar multiplications, improving on the 49","baseline":"49 (using two recursion levels of Strassen's algorithm)","result":"48","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://arxiv.org/abs/2506.13242","https://deepmind.google/discover/blog/alphaevolve-a-gemini-powered-coding-agent-for-designing-advanced-algorithms/","https://arxiv.org/abs/2603.18699","https://arxiv.org/abs/2602.13171","https://journals.rcsi.science/2686-9543/article/view/269374","https://arxiv.org/abs/2607.15069","https://arxiv.org/pdf/2506.13131"],"source_refs":["AIBD-0076-source-0001","AIBD-0076-source-0002","AIBD-0076-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the previously best known algorithm for 4x4 matrices over the complex numbers, two recursion levels of Strassen's 1969 algorithm","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 49 (using two recursion levels of Strassen's algorithm). Result: 48.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"4x4 complex-valued matrices with 48 scalar multiplications, improving on the 49","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0076-reported-measure","evaluation_data":null},{"case_id":"AIBD-0130","case_url":"records/AIBD-0130.html","measure":null,"reported_result":null,"baseline":null,"result":"λίγα δευτερόλεπτα πριν σημειωθεί η σεισμική δόνηση","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.newsbomb.gr/technologia/story/1649410/seismos-tora-pos-na-energopoiisete-ti-nea-efarmogi-tis-google-pou-enimeronei-gia-seismoys","https://www.gizchina.gr/?p=363710","https://www.in.gr/2025/05/15/in-science/technology/pos-tha-energopoiisete-tin-efarmogi-tis-google-pou-proeidopoiei-gia-seismo","https://www.alphafreepress.gr/2025/05/14/ellada/kasos-seismos-61-rixter-pos-leitourgei-to-systima-proeidopoiisis-sta-android","https://earthquake.usgs.gov/earthquakes/eventpage/us7000pz2p","https://blog.google/products/android/earthquake-detection-and-alerts/","https://crisisresilience.google/android-early-earthquake-warnings","https://doi.org/10.1126/science.ads4779"],"source_refs":["AIBD-0130-source-0001","AIBD-0130-source-0002","AIBD-0130-source-0003","AIBD-0130-source-0004","AIBD-0130-source-0005","AIBD-0130-source-0006","AIBD-0130-source-0007","AIBD-0130-source-0008"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the arrival of the seismic shaking (no warning)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Result: λίγα δευτερόλεπτα πριν σημειωθεί η σεισμική δόνηση.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0130-reported-measure","evaluation_data":null},{"case_id":"AIBD-0131","case_url":"records/AIBD-0131.html","measure":"notifications had reached their phones 30 seconds before the quake","reported_result":"notifications had reached their phones 30 seconds before the quake","baseline":null,"result":"Bu deprem olmadan 30 saniye öncesinde telefonlarımıza bildirim geldi","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.haberler.com/guncel/karadeniz-teknik-universitesi-ogrencileri-tbmm-de-depremi-hissetti-18634729-haberi","https://www.yeniakit.com.tr/haber/milletvekilleri-deprem-konusurken-depreme-yakalandi-tbmmde-deprem-panigi-1933002.html","https://www.yenicaggazetesi.com/erken-uyar%C4%B1-sistemini-vekillere-anlatirken-depreme-yakalandilar-915433h.htm","https://www.yenicaggazetesi.com/deprem-erken-uyar%C4%B1-sistemini-anlatirken-tbmmde-depreme-yakalandilar-915462h.htm","https://earthquake.usgs.gov/earthquakes/eventpage/us7000pzcl"],"source_refs":["AIBD-0131-source-0001","AIBD-0131-source-0002","AIBD-0131-source-0003","AIBD-0131-source-0004","AIBD-0131-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: feeling the quake with no warning","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Result: Bu deprem olmadan 30 saniye öncesinde telefonlarımıza bildirim geldi.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"notifications had reached their phones 30 seconds before the quake","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0131-reported-measure","evaluation_data":null},{"case_id":"AIBD-0325","case_url":"records/AIBD-0325.html","measure":null,"reported_result":null,"baseline":"no treatment protocol, and the case would die","result":"discharged after 24 hours, by the cousin's account","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.elbalad.news/6583785","https://oecd.ai/en/incidents/2025-05-24-415b","https://www.dostor.org/5081951","https://www.masrawy.com/news/news_regions/details/2025/5/25/2793185/","https://www.dostor.org/5083066","https://www.aljazeeramubasher.net/news/reports/2025/5/26/%d9%82%d8%b1%d8%b5-%d8%a7%d9%84%d8%ba%d9%84%d8%a9-%d9%88%d8%a7%d9%84%d9%85%d9%86%d8%b4%d9%88%d8%b1-%d8%a7%d9%84%d8%b0%d9%8a-%d8%ad%d9%8a%d8%b1","https://openai.com/index/introducing-gpt-5/","https://pmc.ncbi.nlm.nih.gov/articles/PMC10939350/"],"source_refs":["AIBD-0325-source-0002","AIBD-0325-source-0001","AIBD-0325-source-0003","AIBD-0325-source-0004","AIBD-0325-source-0005","AIBD-0325-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0325-reported-measure","evaluation_data":null},{"case_id":"AIBD-0383","case_url":"records/AIBD-0383.html","measure":"downed six Shaheds in six uses","reported_result":"downed six Shaheds in six uses","baseline":"a costly air defense missile spent on each drone","result":"a reusable turret that costs about $150,000, against jet-powered drones that cost $100,000 to $200,000 each","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://united24media.com/war-in-ukraine/ukraines-new-ai-powered-turret-sky-sentinel-is-already-shooting-down-russias-fastest-drones-23750","https://united24media.com/war-in-ukraine/ai-powered-turret-that-hunts-russian-drones-meet-sky-sentinel-ukraines-new-air-defense-8589","https://euromaidanpress.com/2025/05/28/forbes-ukraine-deploys-ai-turrets-against-record-russian-drone-attacks/","https://thedefensepost.com/2025/06/03/ukraine-ai-shahed-drones","https://oecd.ai/en/incidents/2025-06-02-0176","https://www.businessinsider.com/ukraine-ai-powered-turret-shoot-russia-shahed-drones-sky-sentinel-2025-6","https://kyivindependent.com/new-automatic-gun-turrets-successfully-down-first-russian-jet-powered-drones-air-force-says/"],"source_refs":["AIBD-0383-source-0001","AIBD-0383-source-0002","AIBD-0383-source-0003","AIBD-0383-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0383-reported-measure","evaluation_data":null},{"case_id":"AIBD-0632","case_url":"records/AIBD-0632.html","measure":"Time-dependent AUC at five years and calibration (modified GND test).","reported_result":"an AUC of 0.70, above a goal of 0.64","baseline":"AUC goal of 0.64","result":"AUC 0.70 (95% CI 0.69 to 0.72)","baseline_value":null,"result_value":null,"unit":null,"effect":"AUC 0.70 overall; 0.66 excluding current cancers; calibration endpoints met.","design":"other.","sample":"77,511 exams from 44,112 patients at five US screening sites not used for training.","uncertainty":"Retrospective; no same-sample comparison with traditional models; overestimation in the highest decile for patients without current cancer.","comparison_basis":"assumed","source_urls":["https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN240047.pdf"],"source_refs":["AIBD-0632-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Fixed performance goal of 0.64.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"AUC 0.70 overall; 0.66 excluding current cancers; calibration endpoints met.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One screening mammogram per exam.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective validation of five-year risk predictions against biopsy-proven cancer and follow-up mammography, adjusted for censoring.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time-dependent AUC at five years and calibration (modified GND test).","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"None reported.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"77,511 exams from 44,112 patients at five US screening sites not used for training.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective; no same-sample comparison with traditional models; overestimation in the highest decile for patients without current cancer.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0632-reported-measure","evaluation_data":null},{"case_id":"AIBD-0133","case_url":"records/AIBD-0133.html","measure":null,"reported_result":null,"baseline":"the car would slow and stop within the current lane of the highway","result":"the car slowed, moved to the rightmost emergency lane, turned on its hazard lights and placed an SOS call","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://cnevpost.com/2025/05/30/nio-rolls-out-first-version-nwm/","https://news.mydrivers.com/1/1051/1051320.htm","http://auto.hexun.com/2025-06-03/219371749.html","http://auto.cnfol.com/xincheshangshi/20250603/31358495.shtml","https://news.zol.com.cn/991/9916102.html","https://cnevpost.com/2026/01/28/nio-rolls-out-major-update-to-nwm-driver-assist-china/"],"source_refs":["AIBD-0133-source-0001","AIBD-0133-source-0002","AIBD-0133-source-0003","AIBD-0133-source-0004","AIBD-0133-source-0005","AIBD-0133-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0133-reported-measure","evaluation_data":null},{"case_id":"AIBD-0386","case_url":"records/AIBD-0386.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0386-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0386-reported-measure","evaluation_data":null},{"case_id":"AIBD-0387","case_url":"records/AIBD-0387.html","measure":null,"reported_result":null,"baseline":"Local training on separate simulated nodes","result":"Federated training with FedAvg on the CAFEIN platform, which the makers report as a significant improvement; local training and FedDyn gave the lowest performance, and centralized training gave the highest","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://arxiv.org/html/2410.13869v1","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-25"],"source_refs":["AIBD-0387-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0387-reported-measure","evaluation_data":null},{"case_id":"AIBD-0388","case_url":"records/AIBD-0388.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0388-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0388-reported-measure","evaluation_data":null},{"case_id":"AIBD-0390","case_url":"records/AIBD-0390.html","measure":"Study 1: Marginal and internal gaps and occlusal discrepancies. Study 2: Design time and occlusal surface accuracy.","reported_result":"Study 1: Fit was equivalent at the buccal, mesial and distal margins and internally; Dentbird crowns had larger lingual margin gaps and larger occlusal discrepancies. Study 2: Design time fell from 397.2 to 99.4 min for the premolar and from 516.4 to 97.6 min for the molar. The occlusal deviation between technicians' crowns fell from 275.5 to 25.7 µm.","baseline":"397.2 ± 80.4 min for the premolar crown with conventional CAD","result":"99.4 ± 17.1 min for the premolar crown with Dentbird","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: Fit was equivalent at the buccal, mesial and distal margins and internally; Dentbird crowns had larger lingual margin gaps and larger occlusal discrepancies. Study 2: Design time fell from 397.2 to 99.4 min for the premolar and from 516.4 to 97.6 min for the molar. The occlusal deviation between technicians' crowns fell from 275.5 to 25.7 µm.","design":"Study 1: controlled observational. Study 2: controlled observational.","sample":"Study 1: 62 patients needing complete crowns on premolars and molars, Korea. Study 2: Five technicians and two crowns each on one master model, Kanagawa Dental University, Japan.","uncertainty":"Study 1: Interim resin crowns; one technician as the comparator; no measure of design time or patient-reported outcomes. Study 2: One model, five technicians and twenty crowns; a single measurer; the reported design times are long for a single crown and the paper does not explain how time was counted.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11886404/","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-43","https://www.sciencedirect.com/science/article/abs/pii/S0300571224003117","https://pmc.ncbi.nlm.nih.gov/articles/PMC12411301/","https://pubmed.ncbi.nlm.nih.gov/40234164/","https://pubmed.ncbi.nlm.nih.gov/41353038/"],"source_refs":["AIBD-0390-source-0004","AIBD-0390-source-0003"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 2: About four to five times faster design with Dentbird.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Crowns designed by a technician with more than 10 years of CAD experience in conventional software. Study 2: 3Shape Dental System used by the same technicians.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: controlled observational. Study 2: controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Fit was equivalent at the buccal, mesial and distal margins and internally; Dentbird crowns had larger lingual margin gaps and larger occlusal discrepancies. Study 2: Design time fell from 397.2 to 99.4 min for the premolar and from 516.4 to 97.6 min for the molar. The occlusal deviation between technicians' crowns fell from 275.5 to 25.7 µm.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: controlled observational. Study 2: controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Each patient received one crown designed by Dentbird and one designed by a technician in conventional CAD for the same prepared tooth; fit was measured by a digital triple-scan technique. Study 2: Five technicians each designed two crowns with conventional CAD and with Dentbird on the same model; design time, occlusal accuracy and marginal fit were measured by one dentist.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Marginal and internal gaps and occlusal discrepancies. Study 2: Design time and occlusal surface accuracy.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 62 patients needing complete crowns on premolars and molars, Korea. Study 2: Five technicians and two crowns each on one master model, Kanagawa Dental University, Japan.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Interim resin crowns; one technician as the comparator; no measure of design time or patient-reported outcomes. Study 2: One model, five technicians and twenty crowns; a single measurer; the reported design times are long for a single crown and the paper does not explain how time was counted.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0390-reported-measure","evaluation_data":null},{"case_id":"AIBD-0391","case_url":"records/AIBD-0391.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0391-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0391-reported-measure","evaluation_data":null},{"case_id":"AIBD-0397","case_url":"records/AIBD-0397.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0397-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0397-reported-measure","evaluation_data":null},{"case_id":"AIBD-0399","case_url":"records/AIBD-0399.html","measure":"Study 1: History-taking assessment score. Study 2: Human ranking score and agreement with scientific consensus.","reported_result":"Study 1: 10.50-point increase (95% CI 4.66 to 16.33, p < 0.001). Study 2: Ranking score 0.60 against 0.48 for the baseline and 0.61 for GPT-4; scientific consensus 84.0 percent against 46.5 percent for the baseline.","baseline":"control group score on the history-taking assessment","result":"10.50 points higher (95% CI 4.66 to 16.33) for students trained with the digital-patient system","baseline_value":null,"result_value":null,"unit":null,"effect":"Study 1: 10.50-point increase (95% CI 4.66 to 16.33, p < 0.001). Study 2: Ranking score 0.60 against 0.48 for the baseline and 0.61 for GPT-4; scientific consensus 84.0 percent against 46.5 percent for the baseline.","design":"Study 1: randomized. Study 2: other.","sample":"Study 1: 84 fourth-year clinical medicine students at the Zhongshan School of Medicine, Sun Yat-sen University; single center; registered as NCT06229379. Study 2: ChatZOC and 10 other language models, including GPT-4 and GPT-3.5 Turbo; no patients involved.","uncertainty":"Study 1: Single center, small sample and author team from the developer; the outcome is a test score and does not measure later care of patients. Study 2: A benchmark by the developers without clinical patients; the authors state that real-world use needs further study.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12322286/","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-97","https://jamanetwork.com/journals/jamaophthalmology/fullarticle/2820953"],"source_refs":["AIBD-0399-source-0003","https://pmc.ncbi.nlm.nih.gov/articles/PMC11258636/"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: The trial compares training with the LLM digital patient against training without it, so the difference reflects the system as used in the course. Study 2: The comparison isolates the effect of adding the ophthalmic knowledge base to the baseline model.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Students trained by traditional methods. Study 2: The baseline model without retrieval, and GPT-4.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: randomized. Study 2: other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 10.50-point increase (95% CI 4.66 to 16.33, p < 0.001). Study 2: Ranking score 0.60 against 0.48 for the baseline and 0.61 for GPT-4; scientific consensus 84.0 percent against 46.5 percent for the baseline.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Training sessions with voice-enabled virtual patients built from de-identified records.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: randomized. Study 2: other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Standardized medical history-taking assessment scores after training. Study 2: Double-masked expert ranking of answers to 300 ophthalmology clinical questions.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: History-taking assessment score. Study 2: Human ranking score and agreement with scientific consensus.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Code for the LLMDP system is posted on GitHub (pangjianyu-sunivers/LLMDP).","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: 84 fourth-year clinical medicine students at the Zhongshan School of Medicine, Sun Yat-sen University; single center; registered as NCT06229379. Study 2: ChatZOC and 10 other language models, including GPT-4 and GPT-3.5 Turbo; no patients involved.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Study 1: Single center, small sample and author team from the developer; the outcome is a test score and does not measure later care of patients. Study 2: A benchmark by the developers without clinical patients; the authors state that real-world use needs further study.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0399-reported-measure","evaluation_data":null},{"case_id":"AIBD-0400","case_url":"records/AIBD-0400.html","measure":null,"reported_result":null,"baseline":"Traditional methods (no figure given)","result":"30 to 50 times the efficiency and a 300% gain in accuracy","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.yunnanbaiyao.com.cn/view/ynbyPc/1/92/view/7804.html"],"source_refs":["AIBD-0400-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0400-reported-measure","evaluation_data":null},{"case_id":"AIBD-0412","case_url":"records/AIBD-0412.html","measure":null,"reported_result":null,"baseline":"40 seconds per image by a person","result":"40 milliseconds per image by GeoGPT","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-176","https://www.journalofgeoethics.eu/index.php/jgsg/article/view/119"],"source_refs":["AIBD-0412-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0412-reported-measure","evaluation_data":null},{"case_id":"AIBD-0413","case_url":"records/AIBD-0413.html","measure":"Average reduction in root mean square error on key variables such as 2-meter temperature and 10-meter wind.","reported_result":"57.8% lower average error than ECMWF-IFS and 47.0% lower than interpolation.","baseline":"ECMWF-IFS forecast error on key regional variables","result":"Baguan-Regional forecast error 57.8% lower on average","baseline_value":null,"result_value":null,"unit":null,"effect":"57.8% lower average error than ECMWF-IFS and 47.0% lower than interpolation.","design":"other.","sample":"China Meteorological Administration Land Data Assimilation System data at 0.05 degree spacing over China for 2022 and 2023; the last half year held out for testing.","uncertainty":"Run by the developers on one half-year test period; no confidence intervals; no independent replication; no measure of downstream grid effects.","comparison_basis":"measured","source_urls":["https://arxiv.org/html/2505.13873v1","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-182","https://web.archive.org/web/2026/https://www.alizila.com/alibaba-damo-academy-ai-model-weather-forecasting/"],"source_refs":["AIBD-0413-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"57.8% average error reduction against the physics-based ECMWF-IFS forecast.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"ECMWF-IFS forecasts and simple interpolation of Baguan's global forecasts.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"57.8% lower average error than ECMWF-IFS and 47.0% lower than interpolation.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Forecasts at 6-hour and 24-hour lead times over China.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective forecast test on held-out data: root mean square error of forecasts compared with gridded analysis data.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Average reduction in root mean square error on key variables such as 2-meter temperature and 10-meter wind.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Regional observation and analysis data are proprietary according to the ITU submission.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"China Meteorological Administration Land Data Assimilation System data at 0.05 degree spacing over China for 2022 and 2023; the last half year held out for testing.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Run by the developers on one half-year test period; no confidence intervals; no independent replication; no measure of downstream grid effects.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0413-reported-measure","evaluation_data":null},{"case_id":"AIBD-0423","case_url":"records/AIBD-0423.html","measure":"Forecast error per variable and lead time, shown as scorecards.","reported_result":"Fine-tuned AuroraSmall improved on almost all variables, most of all up to 48 hours, and beat pretrained AuroraLarge modestly over South Africa; specific humidity did not improve.","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":"Fine-tuned AuroraSmall improved on almost all variables, most of all up to 48 hours, and beat pretrained AuroraLarge modestly over South Africa; specific humidity did not improve.","design":"other.","sample":"South Africa region at 0.25-degree resolution, HRES T0 data for 2022; Europe and the USA as reference regions.","uncertainty":"MSc study described in a blog post without peer review; the authors name the lack of high-resolution local data and limited compute as limits, and accuracy over South Africa stays below Europe and the USA.","comparison_basis":"measured","source_urls":["https://thedataandaiteacher.substack.com/p/ai-to-the-rescue-making-weather-forecasts","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-242","https://ishango.ai/project-umoya/"],"source_refs":["AIBD-0423-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Pretrained AuroraSmall and pretrained AuroraLarge.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Fine-tuned AuroraSmall improved on almost all variables, most of all up to 48 hours, and beat pretrained AuroraLarge modestly over South Africa; specific humidity did not improve.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Hindcast scorecards comparing forecast error of fine-tuned and pretrained Aurora models by variable and lead time.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Forecast error per variable and lead time, shown as scorecards.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"https://github.com/koomited/AIMS-PROJECT.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"South Africa region at 0.25-degree resolution, HRES T0 data for 2022; Europe and the USA as reference regions.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"MSc study described in a blog post without peer review; the authors name the lack of high-resolution local data and limited compute as limits, and accuracy over South Africa stays below Europe and the USA.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0423-reported-measure","evaluation_data":null},{"case_id":"AIBD-0425","case_url":"records/AIBD-0425.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://doi.org/10.1109/COMSNETS63942.2025.10885668"],"source_refs":["AIBD-0425-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0425-reported-measure","evaluation_data":null},{"case_id":"AIBD-0428","case_url":"records/AIBD-0428.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0428-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0428-reported-measure","evaluation_data":null},{"case_id":"AIBD-0433","case_url":"records/AIBD-0433.html","measure":"a 30% improvement in carbon absorption","reported_result":"a 30% improvement in carbon absorption","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-302"],"source_refs":["AIBD-0433-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0433-reported-measure","evaluation_data":null},{"case_id":"AIBD-0439","case_url":"records/AIBD-0439.html","measure":"false-positive and false-negative rates of approximately 6%","reported_result":"false-positive and false-negative rates of approximately 6%","baseline":"manual review that China Telecom calls costly and inefficient at scale (no figure)","result":"automated screening across a network of 622 million users, with false-positive and false-negative rates of about 6% and no figure for cost or speed","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://freedomhouse.org/country/china/freedom-net/2024","https://www.chinalawtranslate.com/en/generative-ai-interim/"],"source_refs":["AIBD-0439-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0439-reported-measure","evaluation_data":null},{"case_id":"AIBD-0440","case_url":"records/AIBD-0440.html","measure":"coverage of quality inspection from 10% to 100%","reported_result":"coverage of quality inspection from 10% to 100%","baseline":"10% coverage by manual inspection","result":"100% coverage by AI agents","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-345","https://mp.weixin.qq.com/s/_DqndMzI7Pb_5tG7WR0CIg"],"source_refs":["AIBD-0440-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0440-reported-measure","evaluation_data":null},{"case_id":"AIBD-0441","case_url":"records/AIBD-0441.html","measure":null,"reported_result":null,"baseline":"Three months of training for newly hired agents","result":"One week of training for newly hired agents","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0441-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0441-reported-measure","evaluation_data":null},{"case_id":"AIBD-0442","case_url":"records/AIBD-0442.html","measure":null,"reported_result":null,"baseline":"10 minutes to detect an anomaly","result":"1 minute to detect an anomaly, with average accuracy above 90%","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-357","http://www.caict.ac.cn/kxyj/qwfb/bps/202409/P020240926401677616677.pdf"],"source_refs":["AIBD-0442-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0442-reported-measure","evaluation_data":null},{"case_id":"AIBD-0443","case_url":"records/AIBD-0443.html","measure":null,"reported_result":null,"baseline":"Representatives clicked through screens and typed customer details into order forms by hand.","result":"The assistant fills order forms from the conversation; the developer reports that key process error rates fell by over 50% and that the average marketing success rate rose by over 30%.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0443-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0443-reported-measure","evaluation_data":null},{"case_id":"AIBD-0448","case_url":"records/AIBD-0448.html","measure":"lowered packet loss and delay at MTN","reported_result":"lowered packet loss and delay at MTN","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-393","https://techafricanews.com/2025/11/13/huawei-and-mtn-south-africa-drive-global-innovation-with-first-an-l4-ip-network-certification/"],"source_refs":["AIBD-0448-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0448-reported-measure","evaluation_data":null},{"case_id":"AIBD-0449","case_url":"records/AIBD-0449.html","measure":"saves more than 900 million kWh a year","reported_result":"saves more than 900 million kWh a year","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-399"],"source_refs":["AIBD-0449-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0449-reported-measure","evaluation_data":null},{"case_id":"AIBD-0451","case_url":"records/AIBD-0451.html","measure":"Macro F1 score.","reported_result":"93.83%.","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":"93.83%.","design":"simulation.","sample":"A test dataset that contains new anomalous traffic flows.","uncertainty":"The authors give no baseline figure, no confidence interval and no field data in the abstract. The authors are the system's developers.","comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://doi.org/10.1109/MSWiM67937.2025.11308786"],"source_refs":["AIBD-0451-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"unresolved","review_state":"reviewed"},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"simulation.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"93.83%.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"simulation.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Test of an online, unsupervised federated model on unlabeled network flow data from 5G core network functions.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Macro F1 score.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"A test dataset that contains new anomalous traffic flows.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The authors give no baseline figure, no confidence interval and no field data in the abstract. The authors are the system's developers.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0451-reported-measure","evaluation_data":null},{"case_id":"AIBD-0453","case_url":"records/AIBD-0453.html","measure":null,"reported_result":null,"baseline":"in the order of several Gbps and Mbps respectively","result":"about a few hundred bytes to a few kilo bytes per frame","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://doi.org/10.1109/COMSNETS63942.2025.10885668"],"source_refs":["AIBD-0453-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0453-reported-measure","evaluation_data":null},{"case_id":"AIBD-0455","case_url":"records/AIBD-0455.html","measure":null,"reported_result":null,"baseline":"The ministry states that manual registration causes delays.","result":"No measured change is reported.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-436"],"source_refs":["AIBD-0455-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0455-reported-measure","evaluation_data":null},{"case_id":"AIBD-0458","case_url":"records/AIBD-0458.html","measure":"gain 300,000 followers and over 210 million views in two months","reported_result":"gain 300,000 followers and over 210 million views in two months","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-454","https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2026-3-PDF-E.pdf"],"source_refs":["AIBD-0458-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0458-reported-measure","evaluation_data":null},{"case_id":"AIBD-0463","case_url":"records/AIBD-0463.html","measure":null,"reported_result":null,"baseline":"The general apps mistranslated construction terms, according to the developer.","result":"Safety Talk reached over 90% translation accuracy in field tests, according to the developer.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://crowdworks.blog/ai-translation-app_itu"],"source_refs":["AIBD-0463-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0463-reported-measure","evaluation_data":null},{"case_id":"AIBD-0465","case_url":"records/AIBD-0465.html","measure":null,"reported_result":null,"baseline":null,"result":"30% less preliminary legal research time for targeted tasks (target)","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0465-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0465-reported-measure","evaluation_data":null},{"case_id":"AIBD-0467","case_url":"records/AIBD-0467.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0467-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0467-reported-measure","evaluation_data":null},{"case_id":"AIBD-0473","case_url":"records/AIBD-0473.html","measure":null,"reported_result":null,"baseline":"floor planning took more than a week","result":"floor planning took roughly one day, a 76% reduction; overall duration fell by 20% for high-density designs","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0473-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0473-reported-measure","evaluation_data":null},{"case_id":"AIBD-0474","case_url":"records/AIBD-0474.html","measure":null,"reported_result":null,"baseline":"dispatchers adjust generation by experience, with decisions taken at the scale of minutes","result":"the agent adjusts generation outputs in real time and, by the developer's account, outperforms dispatchers in control evaluations; decisions can move to the scale of seconds","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0474-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0474-reported-measure","evaluation_data":null},{"case_id":"AIBD-0476","case_url":"records/AIBD-0476.html","measure":null,"reported_result":null,"baseline":"manual inspection, with its stated low efficiency and safety risk","result":"inspection efficiency up by more than 50 percent, manual inspection cost down by 60 percent, and sudden equipment failures down by 40 percent, all stated as expectations","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-568"],"source_refs":["AIBD-0476-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0476-reported-measure","evaluation_data":null},{"case_id":"AIBD-0478","case_url":"records/AIBD-0478.html","measure":null,"reported_result":null,"baseline":"6 person-days","result":"2 person-days","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://web.archive.org/web/2026/https://e.huawei.com/eu/case-studies/industries/manufacturing/202505-gacgroup-digital-large-model"],"source_refs":["AIBD-0478-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0478-reported-measure","evaluation_data":null},{"case_id":"AIBD-0480","case_url":"records/AIBD-0480.html","measure":"Detection accuracy, false alarm rate and missed detection rate; share of edge false positives removed by cloud verification.","reported_result":"Accuracy above 90%, false alarm rate below 5%, missed detection rate below 1%; 92% of edge false positives removed.","baseline":"traditional edge-based camera systems frequently generate false alarms","result":"cloud verification removed 92% of the false positives raised by the edge model","baseline_value":null,"result_value":null,"unit":null,"effect":"Accuracy above 90%, false alarm rate below 5%, missed detection rate below 1%; 92% of edge false positives removed.","design":"other.","sample":"65 videos at Taicang Port: 30 with smoke and 35 without, in daytime, nighttime, rain and fog.","uncertainty":"The developer ran the test and reports bounds only. The sample is small, the 92% figure has no stated base, and the account gives no test dates or effects on safety.","comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0480-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Edge model alone for the 92% figure; none stated for the test rates.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Accuracy above 90%, false alarm rate below 5%, missed detection rate below 1%; 92% of edge false positives removed.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"On-site test of the deployed detector on recorded videos at Taicang Port, run by the developer.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Detection accuracy, false alarm rate and missed detection rate; share of edge false positives removed by cloud verification.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"65 videos at Taicang Port: 30 with smoke and 35 without, in daytime, nighttime, rain and fog.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The developer ran the test and reports bounds only. The sample is small, the 92% figure has no stated base, and the account gives no test dates or effects on safety.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0480-reported-measure","evaluation_data":null},{"case_id":"AIBD-0482","case_url":"records/AIBD-0482.html","measure":null,"reported_result":null,"baseline":"9 minutes 31 seconds of training, 50% accuracy and 66.7% recall with manual training","result":"5 minutes 27 seconds of training, 100% accuracy and 100% recall with the solution chosen by IndustryGPT","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-604","https://www1.hkexnews.hk/app/sehk/2026/108311/documents/sehk26031600232.pdf"],"source_refs":["AIBD-0482-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The reported example compares IndustryGPT-assisted model training with manual training, rather than comparing complete production lines.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0482-reported-measure","evaluation_data":null},{"case_id":"AIBD-0485","case_url":"records/AIBD-0485.html","measure":null,"reported_result":null,"baseline":"customers called the call center with repeated questions and could not find answers in downloaded documents","result":"a bilingual chatbot answers questions and summarizes uploaded documents; the Ministry gives no figures for calls or time saved","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://mof.gov.ae/en/news/ministry-of-finance-launches-digital-transformation-initiatives-using-metaverse-and-ai-solutions/"],"source_refs":["AIBD-0485-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0485-reported-measure","evaluation_data":null},{"case_id":"AIBD-0488","case_url":"records/AIBD-0488.html","measure":null,"reported_result":null,"baseline":"not stated","result":"processing time reduced by up to 95 percent","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-641","https://www.instagram.com/mofuae/reel/DBLmoYbyofv/"],"source_refs":["AIBD-0488-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0488-reported-measure","evaluation_data":null},{"case_id":"AIBD-0490","case_url":"records/AIBD-0490.html","measure":null,"reported_result":null,"baseline":"Traditional manual methods of scrutiny are often time-consuming, resourceintensive, and prone to human error","result":"expected to decrease audit processing time by 50%","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-653"],"source_refs":["AIBD-0490-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0490-reported-measure","evaluation_data":null},{"case_id":"AIBD-0493","case_url":"records/AIBD-0493.html","measure":null,"reported_result":null,"baseline":"manual receipt storage is outdated and time-consuming","result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-671"],"source_refs":["AIBD-0493-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0493-reported-measure","evaluation_data":null},{"case_id":"AIBD-0494","case_url":"records/AIBD-0494.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-677"],"source_refs":["AIBD-0494-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0494-reported-measure","evaluation_data":null},{"case_id":"AIBD-0495","case_url":"records/AIBD-0495.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-683"],"source_refs":["AIBD-0495-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0495-reported-measure","evaluation_data":null},{"case_id":"AIBD-0496","case_url":"records/AIBD-0496.html","measure":null,"reported_result":null,"baseline":"SERA states that no current tool helps people manage and improve their spending patterns.","result":"No measured change is reported.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-689"],"source_refs":["AIBD-0496-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0496-reported-measure","evaluation_data":null},{"case_id":"AIBD-0497","case_url":"records/AIBD-0497.html","measure":null,"reported_result":null,"baseline":"not stated","result":"about 17 percentage points higher on average","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.cherrypot.net/en-US/district/","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-696","https://cherrypot.net/"],"source_refs":["AIBD-0497-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0497-reported-measure","evaluation_data":null},{"case_id":"AIBD-0507","case_url":"records/AIBD-0507.html","measure":null,"reported_result":null,"baseline":"Parents searched fragmented, static listings with no personal filters.","result":"Parents ask questions in plain language and receive KHDA school data; the account reports no measured time saving.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"source_refs":["AIBD-0507-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0507-reported-measure","evaluation_data":null},{"case_id":"AIBD-0509","case_url":"records/AIBD-0509.html","measure":null,"reported_result":null,"baseline":null,"result":"25% improvement in urban management efficiency","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-768"],"source_refs":["AIBD-0509-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0509-reported-measure","evaluation_data":null},{"case_id":"AIBD-0511","case_url":"records/AIBD-0511.html","measure":null,"reported_result":null,"baseline":"About 100 parking spaces","result":"About 50% more parking spaces, with a stated annual return on investment of 30%","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-780"],"source_refs":["AIBD-0511-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0511-reported-measure","evaluation_data":null},{"case_id":"AIBD-0513","case_url":"records/AIBD-0513.html","measure":null,"reported_result":null,"baseline":"fixed, rarely updated content and a lack of interactivity","result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-1-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-793"],"source_refs":["AIBD-0513-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0513-reported-measure","evaluation_data":null},{"case_id":"AIBD-0517","case_url":"records/AIBD-0517.html","measure":null,"reported_result":null,"baseline":"general availability or speed test applications that provide one-off measurements","result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-1-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-817"],"source_refs":["AIBD-0517-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0517-reported-measure","evaluation_data":null},{"case_id":"AIBD-0518","case_url":"records/AIBD-0518.html","measure":null,"reported_result":null,"baseline":"manual, static, and lacks customization based on user needs","result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-1-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-823"],"source_refs":["AIBD-0518-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0518-reported-measure","evaluation_data":null},{"case_id":"AIBD-0520","case_url":"records/AIBD-0520.html","measure":"Accuracy and F1 score.","reported_result":"The authors report that mini-Xception gave the most even scores across datasets, with an F1 score of 0.91 on the custom dataset. They report that the LiteRT phone versions of mini-Xception and YoloClassifier fell to near chance-level accuracy, about 0.50.","baseline":"struggle to accurately detect drowsiness leading to false negatives","result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":"The authors report that mini-Xception gave the most even scores across datasets, with an F1 score of 0.91 on the custom dataset. They report that the LiteRT phone versions of mini-Xception and YoloClassifier fell to near chance-level accuracy, about 0.50.","design":"other.","sample":"Eight convolutional and transformer models tested on public drowsiness datasets and a custom dataset of 16,788 images assembled from public sources.","uncertainty":"Tests used single images. The authors state that the custom dataset represents racial and gender groups equally, but they report scores per dataset and give no scores by group. No test took place in a vehicle or with drivers.","comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-1-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-836","https://zenodo.org/records/21512007","https://zenodo.org/records/20595333/files/D4.5%20Initial%20use%20case%20definition%20and%20AutoML%20validation.pdf"],"source_refs":["AIBD-0520-source-0002"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Published drowsiness models tested across datasets other than those they were trained on.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"The authors report that mini-Xception gave the most even scores across datasets, with an F1 score of 0.91 on the custom dataset. They report that the LiteRT phone versions of mini-Xception and YoloClassifier fell to near chance-level accuracy, about 0.50.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Offline accuracy and F1 scores on held-out images, with and without face cropping and contrast equalization, and after conversion to the LiteRT format for phones.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Accuracy and F1 score.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Eight convolutional and transformer models tested on public drowsiness datasets and a custom dataset of 16,788 images assembled from public sources.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Tests used single images. The authors state that the custom dataset represents racial and gender groups equally, but they report scores per dataset and give no scores by group. No test took place in a vehicle or with drivers.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0520-reported-measure","evaluation_data":null},{"case_id":"AIBD-0527","case_url":"records/AIBD-0527.html","measure":null,"reported_result":null,"baseline":"Decisions from one sensor at a time","result":"Decisions from several diverse sensors combined, with better decision accuracy in a laboratory test","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://arxiv.org/abs/2403.08984","https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://cordis.europa.eu/project/id/101070028/reporting"],"source_refs":["AIBD-0527-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0527-reported-measure","evaluation_data":null},{"case_id":"AIBD-0530","case_url":"records/AIBD-0530.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-903"],"source_refs":["AIBD-0530-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0530-reported-measure","evaluation_data":null},{"case_id":"AIBD-0531","case_url":"records/AIBD-0531.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-909","https://doi.org/10.1109/percomworkshops68308.2026.11585407","https://arxiv.org/abs/2402.17689"],"source_refs":["AIBD-0531-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0531-reported-measure","evaluation_data":null},{"case_id":"AIBD-0532","case_url":"records/AIBD-0532.html","measure":null,"reported_result":null,"baseline":"Configuring a scenario typically takes several minutes and needs technical knowledge.","result":"The makers state that a voice request completes the task in seconds; no timing study is reported.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://raw.githubusercontent.com/Chongqing-Changan-Automobile-Co-Ltd/AI_for_good/main/Changan_Qiyuan.pdf"],"source_refs":["AIBD-0532-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0532-reported-measure","evaluation_data":null},{"case_id":"AIBD-0533","case_url":"records/AIBD-0533.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-921","https://optmsol.com/press-releases/elevate-nxt-2026-programme/"],"source_refs":["AIBD-0533-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0533-reported-measure","evaluation_data":null},{"case_id":"AIBD-0534","case_url":"records/AIBD-0534.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-927"],"source_refs":["AIBD-0534-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0534-reported-measure","evaluation_data":null},{"case_id":"AIBD-0535","case_url":"records/AIBD-0535.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-933"],"source_refs":["AIBD-0535-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0535-reported-measure","evaluation_data":null},{"case_id":"AIBD-0536","case_url":"records/AIBD-0536.html","measure":"Battery drop in percent; arrival time; energy efficiency.","reported_result":"Battery drop of 19.8% with AI advice against 27% without it over 62 km, with a 5.6% faster arrival; about 9.2% better energy efficiency from reinforcement-based feedback.","baseline":"Battery drop of 27% over 62km","result":"Battery drop of 19.8% with 5.6% faster ETA","baseline_value":null,"result_value":null,"unit":null,"effect":"Battery drop of 19.8% with AI advice against 27% without it over 62 km, with a 5.6% faster arrival; about 9.2% better energy efficiency from reinforcement-based feedback.","design":"simulation.","sample":"Simulated routes; the team gives one pair of sample results (Route A and Route B) and does not state how many runs it made.","uncertainty":"The team labels the figures as sample results and gives no simulator, run count, variance or route details. It labels the two runs as different routes. The forecasting error figures and the language-model latency that it reports carry citations to outside references.","comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-939"],"source_refs":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Route A with no AI advice.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"simulation.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Battery drop of 19.8% with AI advice against 27% without it over 62 km, with a 5.6% faster arrival; about 9.2% better energy efficiency from reinforcement-based feedback.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Simulated driving only.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"simulation.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Battery use and arrival time compared in a simulated drive with and without AI driving advice.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Battery drop in percent; arrival time; energy efficiency.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Code was to be published after the proof of concept; no repository had appeared by October 2026.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Simulated routes; the team gives one pair of sample results (Route A and Route B) and does not state how many runs it made.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The team labels the figures as sample results and gives no simulator, run count, variance or route details. It labels the two runs as different routes. The forecasting error figures and the language-model latency that it reports carry citations to outside references.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0536-reported-measure","evaluation_data":null},{"case_id":"AIBD-0606","case_url":"records/AIBD-0606.html","measure":"Percent of words in sentences recognized.","reported_result":"bimodal sentence scores on the AzBio test reached 79% with the DNN program against 60%","baseline":"60% sentence recognition in babble (bimodal)","result":"79% sentence recognition in babble (bimodal)","baseline_value":null,"result_value":null,"unit":null,"effect":"79% against 60% in babble (p < 0.001).","design":"controlled observational.","sample":"11 bimodal cochlear implant users aged 71 to 89, Mayo Clinic.","uncertainty":"Eleven patients; speech and babble from one loudspeaker; comparator without any noise reduction; no daily-life measures; the senior author disclosed Sonova consulting or research support in a December 2024 study.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12346913/"],"source_refs":["AIBD-0606-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"19 percentage points.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Calm Situation program without noise reduction.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"79% against 60% in babble (p < 0.001).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"DNN program used during testing.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"AzBio sentence recognition in quiet and in multi-talker babble, implant alone and bimodal.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Percent of words in sentences recognized.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Data in the article.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"11 bimodal cochlear implant users aged 71 to 89, Mayo Clinic.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Eleven patients; speech and babble from one loudspeaker; comparator without any noise reduction; no daily-life measures; the senior author disclosed Sonova consulting or research support in a December 2024 study.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0606-reported-measure","evaluation_data":null},{"case_id":"AIBD-0580","case_url":"records/AIBD-0580.html","measure":null,"reported_result":null,"baseline":"the lifeguard on patrol did not see the man go under","result":"the smartwatch alert gave the lifeguard the swimmer's location with images, and the city reports that it improved his response time","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.stirling.wa.gov.au/city-and-council/news/power-of-anti-drowning-technology-on-display-at-stirling-leisure-%E2%80%93-inglewood-in-wa%E2%80%99s-first-ai-assist","https://7news.com.au/news/ai-pool-monitoring-system-helps-save-life-of-swimmer-suffering-medical-episode-in-the-water-c-19203045","https://www.abc.net.au/news/2025-12-31/ai-system-to-help-lifeguards-in-pools-swim-safety/106172140","https://www.lynxight.com/legal/privacy-policy-pools-he"],"source_refs":["AIBD-0580-source-0001","AIBD-0580-source-0002","AIBD-0580-source-0003","AIBD-0580-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0580-reported-measure","evaluation_data":null},{"case_id":"AIBD-0555","case_url":"records/AIBD-0555.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.gov.uk/government/publications/flood-and-coastal-risk-management-national-report/flood-and-coastal-erosion-risk-management-report-1-april-2024-to-31-march-2025","https://www.local.gov.uk/case-studies/northumberland-county-council-floodai-using-ai-detect-flash-floods","https://www.gov.uk/government/publications/200-million-flood-and-coastal-innovation-programmes-mid-programme-report/200-million-flood-and-coastal-innovation-programmes-mid-programme-report"],"source_refs":["AIBD-0555-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0555-reported-measure","evaluation_data":null},{"case_id":"AIBD-0316","case_url":"records/AIBD-0316.html","measure":"scan over 360 hectares of confirmed and suspected hazardous areas","reported_result":"scan over 360 hectares of confirmed and suspected hazardous areas","baseline":"areas that are too dangerous for manual inspection","result":"survey dangerous areas quickly, precisely, and without risk to people","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://web.archive.org/web/20260114205650/https://www.undp.org/ukraine/press-releases/undp-equips-sesu-cutting-edge-ukrainian-aerial-detection-technology-accelerate-humanitarian-demining-operations","https://www.undp.org/ukraine/press-releases/undp-equips-sesu-cutting-edge-ukrainian-aerial-detection-technology-accelerate-humanitarian-demining-operations"],"source_refs":["AIBD-0316-source-0001","AIBD-0316-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: manual inspection of hazardous areas","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: areas that are too dangerous for manual inspection. Result: survey dangerous areas quickly, precisely, and without risk to people.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"scan over 360 hectares of confirmed and suspected hazardous areas","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0316-reported-measure","evaluation_data":null},{"case_id":"AIBD-0061","case_url":"records/AIBD-0061.html","measure":"colchicine stopped attacks after ~30 years of symptoms","reported_result":"colchicine stopped attacks after ~30 years of symptoms","baseline":"Despite almost 3 decades of symptoms, extensive investigations, including contrast-enhanced abdominal CT and MRI, gastroscopy, colonoscopy, and urological assessment, failed to establish a diagnosis.","result":"the diagnosis of FMF was confirmed","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC13651492/","https://pubmed.ncbi.nlm.nih.gov/41171592/"],"source_refs":["AIBD-0061-source-0001","AIBD-0061-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"almost three decades of conventional investigations","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: Despite almost 3 decades of symptoms, extensive investigations, including contrast-enhanced abdominal CT and MRI, gastroscopy, colonoscopy, and urological assessment, failed to establish a diagnosis. Result: the diagnosis of FMF was confirmed.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"colchicine stopped attacks after ~30 years of symptoms","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0061-reported-measure","evaluation_data":null},{"case_id":"AIBD-0225","case_url":"records/AIBD-0225.html","measure":null,"reported_result":null,"baseline":"a flaw that Google says was known only to threat actors and at risk of being exploited","result":"a fix in SQLite 3.50.2 on 28 June 2025, with no exploitation reported and no listing in the CISA Known Exploited Vulnerabilities catalog","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://blog.google/technology/safety-security/cybersecurity-updates-summer-2025/","https://therecord.media/google-big-sleep-ai-tool-found-bug","https://www.cve.org/CVERecord?id=CVE-2025-6965","https://sqlite.org/releaselog/3_50_2.html","https://sqlite.org/cves.html"],"source_refs":["AIBD-0225-source-0001","AIBD-0225-source-0002","AIBD-0225-source-0003","AIBD-0225-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0225-reported-measure","evaluation_data":null},{"case_id":"AIBD-0187","case_url":"records/AIBD-0187.html","measure":"held $4.41 billion in escrow from about $4.45 billion scheduled for payment","reported_result":"held $4.41 billion in escrow from about $4.45 billion scheduled for payment","baseline":"the approximately $4.45 billion that was scheduled to be paid by Medicare","result":"all but approximately $41 million","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.justice.gov/opa/pr/national-health-care-fraud-takedown-results-324-defendants-charged-connection-over-146","https://www.justice.gov/opa/pr/brooklyn-banker-pleads-guilty-laundering-proceeds-medicare-fraud-transnational-criminal","https://www.justice.gov/criminal/media/1405326/dl?inline=","https://www.justice.gov/opa/pr/ohio-man-pleads-guilty-laundering-health-care-fraud-proceeds-transnational-criminal","https://www.aha.org/news/headline/2024-06-28-cms-releases-proposed-rule-anomalous-dme-spending-and-mssp-financial-calculations"],"source_refs":["AIBD-0187-source-0001","AIBD-0187-source-0002","AIBD-0187-source-0003","AIBD-0187-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the amount Medicare was scheduled to pay on the fraudulent claims had the billing gone undetected","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: the approximately $4.45 billion that was scheduled to be paid by Medicare. Result: all but approximately $41 million.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"HHS-OIG and CMS prevented all but $41 million of $4.45 billion","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0187-reported-measure","evaluation_data":null},{"case_id":"AIBD-0141","case_url":"records/AIBD-0141.html","measure":null,"reported_result":null,"baseline":"For 10+ years I have had several unexplained symptoms.","result":"a few months later and my symptoms are largely resolved","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://api.pullpush.io/reddit/search/submission/?ids=1lrmom4","https://www.reddit.com/r/ChatGPT/comments/1lrmom4/chatgpt_solved_a_10_year_problem_no_doctors_could/","https://www.hindustantimes.com/trending/chatgpt-uncovers-root-cause-of-mysterious-illness-that-baffled-doctors-for-over-10-years-101751704864270.html","https://www.livemint.com/news/trends/reddit-user-claims-chatgpt-uncovered-medical-condition-doctors-overlooked-for-a-decade-11751788742326.html","https://www.pymnts.com/news/artificial-intelligence/2025/chatgpt-as-doctor-when-consumers-rely-on-ai-for-medical-advice/","https://finance.sina.com.cn/jjxw/2025-07-06/doc-infephkm0866335.shtml","https://m.163.com/dy/article/K3SPL1UB0511CSAO.html","https://m.163.com/dy/article/K3PMM22K0511ABV6.html","https://news.ifeng.com/c/8kp8doX5Vd0","https://europepmc.org/article/MED/23288205","https://api.pullpush.io/reddit/search/submission/?ids=1lrmqpj"],"source_refs":["AIBD-0141-source-0001","AIBD-0141-source-0002","AIBD-0141-source-0003","AIBD-0141-source-0004","AIBD-0141-source-0005","AIBD-0141-source-0006","AIBD-0141-source-0007","AIBD-0141-source-0008","AIBD-0141-source-0009"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0141-reported-measure","evaluation_data":null},{"case_id":"AIBD-0147","case_url":"records/AIBD-0147.html","measure":"a fine of up to NT$100,000","reported_result":"a fine of up to NT$100,000","baseline":"After a March 2025 complaint, inspectors found someone burning waste at the site. The bureau told PTS that the site already had cameras, which may have had blind spots, and that dumping had continued","result":"The added system recorded the same man burning waste in July and August 2025. He was identified, confessed and was referred for a fine","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://news.pts.org.tw/article/770398","https://www.epb.taichung.gov.tw/3092215/post","https://www.cna.com.tw/news/asoc/202509130143.aspx","https://news.tvbs.com.tw/local/2988180","http://web.archive.org/web/20250913113400/https://tw.news.yahoo.com/%E7%94%B7%E5%A4%9A%E6%AC%A1%E7%B8%B1%E7%81%AB%E7%87%92%E5%BB%A2%E6%A3%84%E7%89%A9%E5%8D%B1%E5%AE%B3%E5%85%AC%E5%85%B1%E5%AE%89%E5%85%A8-%E6%9C%AA%E9%80%A0%E6%88%90%E5%82%B7%E4%BA%A1-%E5%8F%AA%E8%83%BD%E7%BD%B0%E5%8D%81%E8%90%AC-101000875.html","https://news.ltn.com.tw/news/Taichung/breakingnews/5176646","https://tw.news.yahoo.com/ai%E7%A5%9E%E5%8A%A9%E6%94%BB-%E5%8F%B0%E4%B8%AD%E6%8F%AA%E5%87%BA%E7%84%9A%E7%87%92%E5%BB%A2%E6%A3%84%E7%89%A9%E7%B8%B1%E7%81%AB%E6%85%A3%E7%8A%AF-032546731.html","https://www.ctwant.com/article/445301/","https://www.taichung.gov.tw/8868/8872/9962/3034821","https://www.nownews.com/news/6703456","https://www.taiwannews.com.tw/news/6422182","https://udn.com/news/story/7325/9346776"],"source_refs":["AIBD-0147-source-0001","AIBD-0147-source-0002","AIBD-0147-source-0003","AIBD-0147-source-0004","AIBD-0147-source-0005","AIBD-0147-source-0006","AIBD-0147-source-0007","AIBD-0147-source-0008","AIBD-0147-source-0009","AIBD-0147-source-0010","AIBD-0147-source-0011"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a fine of up to NT$100,000","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a fine of up to NT$100,000","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0147-reported-measure","evaluation_data":null},{"case_id":"AIBD-0281","case_url":"records/AIBD-0281.html","measure":"17 cases involving 19 patients","reported_result":"17 cases involving 19 patients","baseline":null,"result":"changed medical geneticists' testing decisions in 17 cases involving 19 patients","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=EXT_ID:40613735%20AND%20SRC:MED&resultType=core&format=json","https://doi.org/10.1002/ajmg.a.64171","https://europepmc.org/article/MED/40613735","https://pmc.ncbi.nlm.nih.gov/articles/PMC12480590/","https://europepmc.org/article/MED/42661064"],"source_refs":["AIBD-0281-source-0002","AIBD-0281-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the testing strategy the medical geneticists had planned before the software's suggestion","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Result: changed medical geneticists' testing decisions in 17 cases involving 19 patients.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"17 cases involving 19 patients","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0281-reported-measure","evaluation_data":null},{"case_id":"AIBD-0272","case_url":"records/AIBD-0272.html","measure":"11 liver metastases regressed","reported_result":"11 liver metastases regressed","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC13644719/"],"source_refs":["AIBD-0272-source-0001","AIBD-0272-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"11 liver metastases regressed","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"11 liver metastases regressed","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0272-reported-measure","evaluation_data":null},{"case_id":"AIBD-0384","case_url":"records/AIBD-0384.html","measure":"True positive identification rate at rank 1 (the share of searches for people in the database that return the right person first).","reported_result":"100 percent of OIFR probe images returned the correct reference at rank 1, with no difference across demographic groups.","baseline":"a trip to a police station or custody suite to establish identity","result":"identity confirmed on the street from a phone photograph","baseline_value":null,"result_value":null,"unit":null,"effect":"100 percent of OIFR probe images returned the correct reference at rank 1, with no difference across demographic groups.","design":"other.","sample":"Volunteer cohort photographed on the Samsung XcoverPro phone model that South Wales Police used for OIFR, searched against a reference database of 178,400 images; tests run from September to November 2022.","uncertainty":"Processing was retrospective and left out officers taking photographs and judging matches; the evaluator was under contract to the two forces; the test measures accuracy and says nothing about arrests.","comparison_basis":"assumed","source_urls":["https://www.walesonline.co.uk/news/wales-news/man-wanted-2012-arrested-after-32057351","https://emergencyservicestimes.com/2025/07/15/more-than-500-officers-in-wales-now-trained-to-use-real-time-facial-recognition-app/","https://oecd.ai/en/incidents/2025-07-14-7341","https://www.south-wales.police.uk/news/south-wales/news/2025/july/over-500-officers-trained-to-use-new-facial-recognition-mobile-app/","https://ico.org.uk/media2/0hzpqaj5/executive-summary-of-the-south-wales-police-and-gwent-police-audit-report.pdf","https://ico.org.uk/media2/uo1cdoxm/police-forces-frt-outcomes-report-202608.pdf","https://science.police.uk/site/assets/files/3396/frt-equitability-study_mar2023.pdf"],"source_refs":["AIBD-0384-source-0007"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"unresolved","review_state":"reviewed"},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"100 percent of OIFR probe images returned the correct reference at rank 1, with no difference across demographic groups.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Offline batch searches of probe photographs of a volunteer cohort against a reference image database, using a program that NEC supplied to NPL's specification.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"True positive identification rate at rank 1 (the share of searches for people in the database that return the right person first).","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Volunteer cohort photographed on the Samsung XcoverPro phone model that South Wales Police used for OIFR, searched against a reference database of 178,400 images; tests run from September to November 2022.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Processing was retrospective and left out officers taking photographs and judging matches; the evaluator was under contract to the two forces; the test measures accuracy and says nothing about arrests.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0384-reported-measure","evaluation_data":null},{"case_id":"AIBD-0142","case_url":"records/AIBD-0142.html","measure":null,"reported_result":null,"baseline":"My mom had a nonstop cough for 1.5 years. We saw top doctors, visited big hospitals in & out of the city, tried homeopathy, ayurveda, allopathy nothing helped. It got worse: internal bleeding started.","result":"We told a doctor. He confirmed it instantly and changed the meds. Now she’s finally healing.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://in.mashable.com/tech/97585/woman-claims-chatgpt-diagnosed-moms-illness-after-doctors-failed-for-over-a-year","http://web.archive.org/web/20250723122651/https://twitter.com/Life_of_coder/status/1947996833214877981","https://www.ndtv.com/offbeat/chatgpt-saved-my-mom-woman-claims-ai-diagnosed-her-mothers-illness-after-doctors-failed-8940203","https://www.livemint.com/news/trends/chatgpt-saved-my-mom-woman-reveals-how-ai-diagnosed-her-mothers-mysterious-illness-after-1-5-year-struggle-11753346497477.html","https://m.economictimes.com/magazines/panache/top-doctors-failed-for-18-months-then-ai-diagnosed-mystery-illness-woman-shares-how-chatgpt-healed-her-mom/articleshow/122885926.cms","https://timesofindia.indiatimes.com/etimes/trending/chatgpt-saved-my-mom-woman-shares-how-ai-solved-18-month-medical-mystery-that-doctors-missed/articleshow/122881542.cms","https://www.financialexpress.com/life/technology-chatgpt-saved-my-mom-woman-reveals-how-ai-diagnosed-her-mother-s-mysterious-illness-after-1-5-year-struggle-3926225","https://www.freepressjournal.in/tech/chatgpt-saved-my-mom-woman-says-ai-uncovered-correct-cause-of-year-long-cough-doctors-had-deemed-it-fatal","https://cdn.syndication.twimg.com/tweet-result?id=1947996833214877981&token=4pzt91vxhvkzilx","https://cdn.syndication.twimg.com/tweet-result?id=1952293178792440242&token=4pzt91vxhvkzilx"],"source_refs":["AIBD-0142-source-0001","AIBD-0142-source-0002","AIBD-0142-source-0003","AIBD-0142-source-0004","AIBD-0142-source-0005","AIBD-0142-source-0006","AIBD-0142-source-0007","AIBD-0142-source-0008","AIBD-0142-source-0009"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: a year and a half of top doctors, big hospitals and homeopathic, ayurvedic and allopathic treatment","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: My mom had a nonstop cough for 1.5 years. We saw top doctors, visited big hospitals in & out of the city, tried homeopathy, ayurveda, allopathy nothing helped. It got worse: internal bleeding started. Result: We told a doctor. He confirmed it instantly and changed the meds. Now she’s finally healing.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0142-reported-measure","evaluation_data":null},{"case_id":"AIBD-0033","case_url":"records/AIBD-0033.html","measure":"located at about 3,150 m and recovered by 31 July 2025","reported_result":"located at about 3,150 m and recovered by 31 July 2025","baseline":"I metodi di ricerca convenzionali sono rimasti inconcludenti per mesi","result":"meno di 72 ore dalla prima missione del drone al recupero del corpo","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.bluewin.ch/it/attualita/escursionista-scomparso-da-un-anno-ecco-come-e-stato-localizzato-li.2817148","https://www.loscarpone.cai.it/dettaglio/monviso-ritrovato-il-corpo-del-medico-disperso-da-settembre-2024/","https://www.loscarpone.cai.it/dettaglio/sulla-bbc-il-caso-ivaldo-lintelligenza-artificiale-a-supporto-del-soccorso-alpino/","https://www.quadricottero.com/2025/08/droni-e-ia-trovano-il-corpo-del-medico.html","https://www.texsar.org/2025/08/15/adiat-helps-italian-sar-team/","https://www.bbc.com/future/article/20260108-how-ai-solved-the-mystery-of-a-missing-mountaineer","https://www.lastampa.it/cuneo/2025/07/31/news/trovato_il_corpo_del_medico_di_loano_disperso_dal_15_settembre_sul_monviso-15253571/","https://www.wired.it/article/escursionista-disperso-monviso-cadavere-drone-intelligenza-artificiale-soccorso-alpino/","https://www.rainews.it/tgr/piemonte/articoli/2025/07/ritrovato-il-corpo-di-nicola-ivaldo-disperso-sul-monviso-dal-settembre-2024-d14c11b3-3753-454e-98b4-d8919b1648bc.html","https://www.ildolomiti.it/altra-montagna/storie/2026/quel-punto-rosso-era-il-casco-dellalpinista-disperso-da-mesi-senza-lanomalia-segnalata-dallintelligenza-artificiale-in-una-delle-2600-foto-dal-drone-probabilmente-non-sarebbe-mai-stato-trovato","https://torino.corriere.it/notizie/piemonte/25_luglio_31/monviso-ritrovato-dopo-un-anno-e-solo-grazie-ai-droni-il-corpo-di-nicola-ivaldo-medico-savonese-scomparso-in-montagna-c943f224-5e04-4320-856b-a3d4533caxlk.shtml","https://explorersweb.com/ai-solves-the-mystery-of-missing-italian-climber/","https://www.montagna.tv/262960/anche-lintelligenza-artificiale-al-servizio-del-soccorso-alpino/","https://torino.corriere.it/notizie/piemonte/24_settembre_17/disperso-sul-monviso-da-sabato-ha-parcheggiato-l-auto-alla-diga-di-pontechianale-poi-nessuna-traccia-89019eb1-5c94-4928-a627-187960555xlk.shtml","https://oecd.ai/en/incidents/2026-01-10-f4af"],"source_refs":["AIBD-0033-source-0002","AIBD-0033-source-0003","AIBD-0033-source-0004","AIBD-0033-source-0005","AIBD-0033-source-0006","AIBD-0033-source-0007","AIBD-0033-source-0008","AIBD-0033-source-0009","AIBD-0033-source-0001","AIBD-0033-source-0010","AIBD-0033-source-0011","AIBD-0033-source-0012","AIBD-0033-source-0013","AIBD-0033-source-0014","AIBD-0033-source-0015"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"conventional ground and air searches over the preceding months","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: I metodi di ricerca convenzionali sono rimasti inconcludenti per mesi. Result: meno di 72 ore dalla prima missione del drone al recupero del corpo.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"located at about 3,150 m and recovered by 31 July 2025","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0033-reported-measure","evaluation_data":null},{"case_id":"AIBD-0586","case_url":"records/AIBD-0586.html","measure":null,"reported_result":null,"baseline":"The events were known mainly through still photographs and few video records.","result":"The film shows the photographs in motion and adds reconstructed scenes.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.metamorph-award.com/post/bringing-dorothea-lange-s-sealed-photographs-to-life-giovanni-abitante-s-runner-up-mockumentary-i","https://www.persofilmfestival.it/films/i-am-an-american/","https://www.metamorph-award.com/winners-2025"],"source_refs":["AIBD-0586-source-0001","AIBD-0586-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0586-reported-measure","evaluation_data":null},{"case_id":"AIBD-0633","case_url":"records/AIBD-0633.html","measure":"10-year risk of distant metastasis by risk category.","reported_result":"10-year metastasis risk was 28.1% in its high risk category","baseline":"8.1% overall","result":"28.1% in the high risk category; 3.3% in the low risk category","baseline_value":null,"result_value":null,"unit":null,"effect":"High 28.1%, low 3.3%, overall 8.1%.","design":"other.","sample":"886 patients at three US sites.","uncertainty":"No comparison of decisions with and without the software; few African American patients (72).","comparison_basis":"measured","source_urls":["https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN240068.pdf"],"source_refs":["AIBD-0633-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Overall cohort risk.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"High 28.1%, low 3.3%, overall 8.1%.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One biopsy slide set per patient.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Prognostic validation of risk categories against observed 10-year distant metastasis.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"10-year risk of distant metastasis by risk category.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"None reported.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"886 patients at three US sites.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"No comparison of decisions with and without the software; few African American patients (72).","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0633-reported-measure","evaluation_data":null},{"case_id":"AIBD-0064","case_url":"records/AIBD-0064.html","measure":"sensitivity for anterior large vessel occlusion; estimated door-in-door-out saving; CTA-to-thrombectomy notification time.","reported_result":"transferred in 25 minutes for successful thrombectomy","baseline":"2·3% (376 of 15 969 patients)","result":"4·6% (751 of 15 428 patients), a relative increase of 100%","baseline_value":null,"result_value":null,"unit":null,"effect":"sensitivity 83.5% (95% CI 76.5 to 90.5); estimated median maximal DIDO saving 30 minutes (IQR 4 to 45); notification time 11 minutes with routine AI against 48 minutes at the primary centers.","design":"simulation.","sample":"109 patients transferred for thrombectomy from five low-volume primary stroke centers in Poland over 53 months, and 69 patients at a comprehensive stroke center.","uncertainty":"Retrospective design, patients in the 6 to 24 hour window excluded, varying scan protocols, one region of Poland; the time saving is an estimate and the notification comparison is between different kinds of center.","comparison_basis":"measured","source_urls":["https://europepmc.org/abstract/MED/41339157","https://www.england.nhs.uk/2025/12/life-changing-ai-support-helping-stroke-patients-get-a-second-chance/","https://doi.org/10.5603/pjnns.106121","https://www.nice.org.uk/guidance/htg708/chapter/1-Recommendations","https://pmc.ncbi.nlm.nih.gov/articles/PMC11732836/"],"source_refs":["AIBD-0064-source-0004"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"the same patients' actual transfer workflow without the software.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"simulation.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"sensitivity 83.5% (95% CI 76.5 to 90.5); estimated median maximal DIDO saving 30 minutes (IQR 4 to 45); notification time 11 minutes with routine AI against 48 minutes at the primary centers.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"simulation.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Brainomix 360 was run on the original scans of past transfers, and two blinded vascular neurologists simulated transfer decisions; workflow times were compared with a center that used the software routinely.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"sensitivity for anterior large vessel occlusion; estimated door-in-door-out saving; CTA-to-thrombectomy notification time.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"109 patients transferred for thrombectomy from five low-volume primary stroke centers in Poland over 53 months, and 69 patients at a comprehensive stroke center.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective design, patients in the 6 to 24 hour window excluded, varying scan protocols, one region of Poland; the time saving is an estimate and the notification comparison is between different kinds of center.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0064-reported-measure","evaluation_data":null},{"case_id":"AIBD-0305","case_url":"records/AIBD-0305.html","measure":null,"reported_result":null,"baseline":"she was told she was anxious, depressed, had epilepsy","result":"Genetic testing then confirmed the AI suggestion.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.bbc.com/news/articles/cx24njzzkgjo","https://www.walesonline.co.uk/news/wales-news/cardiff-woman-uses-chat-gpt-33722339","https://www.bbc.co.uk/news/articles/c5gmr3dl2lp4o"],"source_refs":["AIBD-0305-source-0001","AIBD-0305-source-0002","AIBD-0305-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: four years of diagnoses by doctors (anxiety, depression, epilepsy)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: she was told she was anxious, depressed, had epilepsy. Result: Genetic testing then confirmed the AI suggestion.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0305-reported-measure","evaluation_data":null},{"case_id":"AIBD-0514","case_url":"records/AIBD-0514.html","measure":"Accuracy.","reported_result":"The developers state that OneAstronomy outperformed all three; the figures appear only in charts.","baseline":"Accuracy of the comparison models on GPQA-astro (447 questions) and the NAOC-TianYi test set; the account gives the figures only in charts","result":"Higher accuracy for OneAstronomy on both test sets, as stated by the developers","baseline_value":null,"result_value":null,"unit":null,"effect":"The developers state that OneAstronomy outperformed all three; the figures appear only in charts.","design":"other.","sample":"GPQA-astro, an astronomy subset of 447 questions, and the NAOC-TianYi test set.","uncertainty":"The developers report the results themselves and name no outside group that ran the tests; they give no confidence intervals, prompts or run details.","comparison_basis":"measured","source_urls":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf","https://www.itu.int/epublications/publication/ai-for-good-innovate-for-impact#_idParaDest-799"],"source_refs":["https://www.itu.int/dms_pub/itu-t/opb/ai4g/T-AI4G-AI4GOOD-2025-5-PDF-E.pdf"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Gain from domain fine-tuning over the Qwen2.5 base model, size not given in the text.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Qwen2.5 base model, GPT-4o and AstroLLaMA 70B.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"The developers state that OneAstronomy outperformed all three; the figures appear only in charts.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Benchmark accuracy, the share of questions answered correctly, run by the developers.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Accuracy.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"GPQA-astro, an astronomy subset of 447 questions, and the NAOC-TianYi test set.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The developers report the results themselves and name no outside group that ran the tests; they give no confidence intervals, prompts or run details.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0514-reported-measure","evaluation_data":null},{"case_id":"AIBD-0134","case_url":"records/AIBD-0134.html","measure":"surgeons removed about two thirds of his pancreas and his symptoms went","reported_result":"surgeons removed about two thirds of his pancreas and his symptoms went","baseline":"Tras varias pruebas le “dijeron que no era nada”","result":"en pocos minutos su diagnóstico señalaba como posible principal causa de lo que me sucedía: un tumor de páncreas, un insulinoma","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.marca.com/tecnologia/2025/08/05/gracias-chatgpt-descubre-cancer-pancreas-marcado-despues-vidas.html","https://www.instagram.com/reel/DM508s_tAKz/","https://www.elespanol.com/ciencia/20250805/enfermo-cancer-pancreas-anos-rotundo-chatgpt-detecto-tumor-medicos/1003743876208_0.html","https://www.telecinco.es/noticias/salud/20250805/chatgpt-ayuda-descubrir-cancer-pancreas-anos-sin-diagnostico_18_016301459.html","https://www.cope.es/programas/herrera-en-cope/audios/descubre-tumor-raro-gracias-ia-meses-diagnostico-medico-sirve-leccion-otros-pacientes-20250807_3196603.html","https://www.abc.es/play/television/noticias/ia-pronostico-posible-diagnostico-cancer-pancreas-acerto-20250808193843-nt.html"],"source_refs":["AIBD-0134-source-0002","AIBD-0134-source-0003","AIBD-0134-source-0004","AIBD-0134-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: three years of visits to specialists, including a neurologist, and a primary-care doctor","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: Tras varias pruebas le “dijeron que no era nada”. Result: en pocos minutos su diagnóstico señalaba como posible principal causa de lo que me sucedía: un tumor de páncreas, un insulinoma.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"surgeons removed about two thirds of his pancreas and his symptoms went","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0134-reported-measure","evaluation_data":null},{"case_id":"AIBD-0135","case_url":"records/AIBD-0135.html","measure":null,"reported_result":null,"baseline":"no person awake in the home to notice her repeated words","result":"the robot flagged the words, the control center phoned her and reported to police, and she was admitted to hospital the next afternoon","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.khan.co.kr/article/202508132034005","https://www.daejeon.go.kr/drh/depart/board/boardNormalView.do?boardId=blog_0001&menuSeq=1625&ntatcSeq=1494637456&pageIndex=1","https://www.yna.co.kr/view/AKR20250813050400063","https://www.hankyung.com/article/2025081394957","https://news.mt.co.kr/mtview.php?no=2025081321115894713","https://www.donga.com/news/Society/article/all/20250813/132180422/2","https://www.yonhapnewstv.co.kr/news/AKR20250813142450392","https://zdnet.co.kr/view/?no=20220405104411","https://www.yna.co.kr/view/AKR20260731144200063","https://www.daejonilbo.com/news/articleView.html?idxno=2226141"],"source_refs":["AIBD-0135-source-0002","AIBD-0135-source-0003","AIBD-0135-source-0004","AIBD-0135-source-0005","AIBD-0135-source-0006","AIBD-0135-source-0007","AIBD-0135-source-0008","AIBD-0135-source-0009"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0135-reported-measure","evaluation_data":null},{"case_id":"AIBD-0178","case_url":"records/AIBD-0178.html","measure":"disconnected more than 82 lakh connections through ASTR after they failed re-verification","reported_result":"disconnected more than 82 lakh connections through ASTR after they failed re-verification","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.pib.gov.in/PressReleasePage.aspx?PRID=2153524","https://www.pib.gov.in/PressReleasePage.aspx?PRID=1922542","https://www.thehindu.com/news/national/17000-extra-sim-cards-of-subscribers-beyond-9-connections-deactivated-in-bihar-and-jharkhand/article66903692.ece","https://oecd.ai/en/incidents/2023-05-28-a3cc","https://www.thehansindia.com/business/govt-disconnects-over-82-lakh-fraudulent-mobile-connections-via-astr-1130704","https://www.medianama.com/2023/05/223-rti-contract-telecom-department-facial-recognition-astr-confidential/","https://www.medianama.com/2026/08/223-dot-nine-sim-limit-subscriber-photos/"],"source_refs":["AIBD-0178-source-0001","AIBD-0178-source-0002","AIBD-0178-source-0003","AIBD-0178-source-0004","AIBD-0178-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"disconnected more than 82 lakh connections through ASTR after they failed re-verification","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"disconnected more than 82 lakh connections through ASTR after they failed re-verification","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0178-reported-measure","evaluation_data":null},{"case_id":"AIBD-0238","case_url":"records/AIBD-0238.html","measure":"share of tasks solved under full validation, compared with the share of original PoCs that no longer crash.","reported_result":"found 54 synthetic vulnerabilities, patched 43, and found 18 real bugs for disclosure","baseline":"37% at semifinals","result":"teams identified 86% of the competition's synthetic vulnerabilities","baseline_value":null,"result_value":null,"unit":null,"effect":"Atlantis passed 92.0% of PoCs and solved 48.4% of tasks; Buttercup and RoboDuck solved 42.7% and 29.6%; PoC-only validation inflated solve rates by 1.83 times on average.","design":"other.","sample":"213 tasks across 32 open-source projects; the AIxCC systems Atlantis (Team Atlanta), Buttercup (Trail of Bits) and RoboDuck (Theori), and eight general-purpose agents.","uncertainty":"C/C++ tasks only. The authors ran the AIxCC systems with models of their choice (GPT-5.6 Sol, and also Claude Opus 4.8 for Atlantis). Atlantis had a larger budget and is not directly comparable. The benchmark measures patching and does not measure discovery.","comparison_basis":"measured","source_urls":["https://www.darpa.mil/news/2025/aixcc-results","https://therecord.media/darpa-ai-code-competition-winner-def-con","https://daniel.haxx.se/blog/2025/10/22/aixcc-curl-details/","https://arxiv.org/abs/2609.04075","https://arxiv.org/abs/2603.08566"],"source_refs":["https://arxiv.org/html/2609.04075v1"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"the same agents judged by original-PoC validation only, and general-purpose agents such as Codex and Claude Code.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Atlantis passed 92.0% of PoCs and solved 48.4% of tasks; Buttercup and RoboDuck solved 42.7% and 29.6%; PoC-only validation inflated solve rates by 1.83 times on average.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Benchmark run of 11 patching agents on 213 C/C++ vulnerabilities transplanted or mutated into new code, with patches checked by extra fuzzing PoCs, sanitizer regression, program output state and unit tests.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"share of tasks solved under full validation, compared with the share of original PoCs that no longer crash.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"https://github.com/ai-sec-lab/PatchBench.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"213 tasks across 32 open-source projects; the AIxCC systems Atlantis (Team Atlanta), Buttercup (Trail of Bits) and RoboDuck (Theori), and eight general-purpose agents.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"C/C++ tasks only. The authors ran the AIxCC systems with models of their choice (GPT-5.6 Sol, and also Claude Opus 4.8 for Atlantis). Atlantis had a larger budget and is not directly comparable. The benchmark measures patching and does not measure discovery.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0238-reported-measure","evaluation_data":null},{"case_id":"AIBD-0136","case_url":"records/AIBD-0136.html","measure":null,"reported_result":null,"baseline":"unsure whether she belonged at the emergency room","result":"went to the emergency room, where staff said that she should be there","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.express.co.uk/life-style/health/2103355/woman-bit-spider-asks-ai-chatbot-advice-told-me-go-e","https://www.tiktok.com/@hair.queen.holli/video/7543332854705540365","https://www.dailystar.co.uk/real-life/woman-asks-chat-gpt-whether-she-should-go-to-ae-35834743","https://www.hindustantimes.com/lifestyle/health/woman-credits-chatgpt-for-saving-her-life-after-spider-bite-left-her-arm-numb-it-told-me-to-immediately-go-to-the-er-101756886357629.html","https://economictimes.indiatimes.com/magazines/panache/woman-bitten-by-spider-turns-to-chatgpt-for-help-and-doctors-say-it-saved-her-life/articleshow/123680161.cms","https://pmc.ncbi.nlm.nih.gov/articles/PMC12920132/"],"source_refs":["AIBD-0136-source-0001","AIBD-0136-source-0002","AIBD-0136-source-0003","AIBD-0136-source-0004","AIBD-0136-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0136-reported-measure","evaluation_data":null},{"case_id":"AIBD-0329","case_url":"records/AIBD-0329.html","measure":null,"reported_result":null,"baseline":null,"result":"the fire was contained before it spread out of control","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.wings-ict-solutions.eu/early-wildfire-detection-in-trizina-by-wings-wi-breathe-innovation-in-action/","https://www.wings-ict-solutions.eu/wings-ict-solutions-delivers-wi-breathe-for-early-wildfire-detection-and-enhanced-environmental-intelligence-in-troizinia-methana-poros/","https://www.ertnews.gr/roi-idiseon/fotia-tora-pyrkagia-sto-diavologefyro-troizinias-megali-kinitopoiisi-epigeion-kai-enaerion-dynameon/","https://www.saronicmagazine.gr/saronikos/troizina/95666-o-dimos-troizinias-methanon-gia-tin-fotia-stin-troizina","https://www.capital.gr/epikairotita/3990977/wings-ict-solutions-me-texnologia-ai-kai-5g-i-puranixneusi-se-perioxes-tou-dimou-troizinias-methanon/"],"source_refs":["AIBD-0329-source-0001","AIBD-0329-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0329-reported-measure","evaluation_data":null},{"case_id":"AIBD-0237","case_url":"records/AIBD-0237.html","measure":"fixed several hundred bugs as a direct result of reports from AI analyzers","reported_result":"fixed several hundred bugs as a direct result of reports from AI analyzers","baseline":"we always address and fix every warning and complaint they report","result":"over two hundred new potential problems","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://daniel.haxx.se/blog/2025/10/10/a-new-breed-of-analyzers/","https://www.theregister.com/software/2025/10/02/curl-project-swamped-with-ai-slop-finds-not-all-ai-is-bad/588432","https://joshua.hu/retrospective-zeropath-ai-sast-source-code-security-scanners-vulnerability","https://daniel.haxx.se/blog/2025/12/23/a-curl-2025-review/"],"source_refs":["AIBD-0237-source-0001","AIBD-0237-source-0002","AIBD-0237-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"curl's existing compilers, static analyzers and fuzzers (scan-build, clang-tidy, CodeSonar, Coverity, CodeQL, OSS-Fuzz), whose findings were already all fixed","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: we always address and fix every warning and complaint they report. Result: over two hundred new potential problems.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"about 50 separately identifiable bugfixes were merged","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0237-reported-measure","evaluation_data":null},{"case_id":"AIBD-0590","case_url":"records/AIBD-0590.html","measure":null,"reported_result":null,"baseline":"In Cohen's example, a catalog search on the Erie Canal faces over 110,000 related items, and general chatbots answer mostly from web sources.","result":"The chatbot passes the request to the library server, which returns works held by the library with direct links.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://newsletter.dancohen.org/archive/the-librarys-new-entryway/","https://newsletter.dancohen.org/archive/the-index-and-the-vector/","https://www.cni.org/topics/digital-libraries/discovery-and-use-reimagined-connecting-scholarly-collections-and-ai-workflows"],"source_refs":["AIBD-0590-source-0001","AIBD-0590-source-0002","AIBD-0590-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0590-reported-measure","evaluation_data":null},{"case_id":"AIBD-0271","case_url":"records/AIBD-0271.html","measure":"Rank of the correct diagnosis among the model's hypotheses.","reported_result":"Oncocytoma ranked first in both cases; the leading diagnosis was wrong for all three published images of other tumor types.","baseline":"Perplexity AI named pyogenic granuloma as the most likely diagnosis and listed oncocytoma among the differentials","result":"GPT-5 named oncocytoma as the leading diagnosis in both cases, matching histology","baseline_value":null,"result_value":null,"unit":null,"effect":"Oncocytoma ranked first in both cases; the leading diagnosis was wrong for all three published images of other tumor types.","design":"case study.","sample":"Two patients with caruncular oncocytoma confirmed by histology at the University of Sassari eye clinic, plus three published images of other caruncular tumors.","uncertainty":"Two cases, one query per case with no repeated runs, no clinician comparison, and selected images.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC13169155/"],"source_refs":["AIBD-0271-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not measured; the authors did not compare the model with clinicians.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Histopathological diagnosis; Perplexity AI given the same images and prompt.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Oncocytoma ranked first in both cases; the leading diagnosis was wrong for all three published images of other tumor types.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"None; the analysis took place after excision and histology.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"case study.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"One standardized prompt per case with slit-lamp photographs plus age, sex and lesion site; the authors recorded whether oncocytoma appeared and its rank among the hypotheses.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Rank of the correct diagnosis among the model's hypotheses.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The prompt text and the slit-lamp figures appear in the article.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two patients with caruncular oncocytoma confirmed by histology at the University of Sassari eye clinic, plus three published images of other caruncular tumors.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Two cases, one query per case with no repeated runs, no clinician comparison, and selected images.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0271-reported-measure","evaluation_data":null},{"case_id":"AIBD-0207","case_url":"records/AIBD-0207.html","measure":"matched a wine thief to six prior thefts and flagged her return","reported_result":"matched a wine thief to six prior thefts and flagged her return","baseline":"staff members had to manually review CCTV footage to identify suspects","result":"staff receive mobile alerts through a corporate app once a flagged person enters any Sheng Siong branch again","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://mothership.sg/2025/08/sheng-siong-face-recognition-cctv/index.html","https://www.channelnewsasia.com/singapore/shoplifter-caught-sheng-siong-facial-recognition-tech-jail-6115421","https://www.straitstimes.com/singapore/courts-crime/woman-who-stole-19-bottles-of-wine-from-sheng-shiong-caught-using-facial-recognition-system","https://www.asiaone.com/singapore/woman-caught-sheng-siongs-facial-recognition-system-stealing-19-bottles-wine-get-8-days"],"source_refs":["AIBD-0207-source-0001","AIBD-0207-source-0002","AIBD-0207-source-0003","AIBD-0207-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the chain's previous practice of staff manually reviewing CCTV footage to identify suspects","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: staff members had to manually review CCTV footage to identify suspects. Result: staff receive mobile alerts through a corporate app once a flagged person enters any Sheng Siong branch again.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"matched a wine thief to six prior thefts and flagged her return","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0207-reported-measure","evaluation_data":null},{"case_id":"AIBD-0662","case_url":"records/AIBD-0662.html","measure":"F1 score of species presence.","reported_result":"BirdNET came close to the experts (F1 0.84)","baseline":"Expert species list of 23 species","result":"BirdNET F1 score of 0.84 against the expert list, with one extra true species","baseline_value":null,"result_value":null,"unit":null,"effect":"F1 0.46 to 0.84 by setting; best 0.84.","design":"controlled observational.","sample":"15.5 hours of rooftop recordings over 60 days at one Munich site.","uncertainty":"One site; short recording scheme; performance depends on settings.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12425287/"],"source_refs":["AIBD-0662-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Near-expert species lists without manual listening.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Expert identification of the same recordings.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"F1 0.46 to 0.84 by setting; best 0.84.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Research use.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Precision, recall and F1 of species lists against expert annotation; linear mixed models of setting effects.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"F1 score of species presence.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Data on Dryad; code on GitHub.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"15.5 hours of rooftop recordings over 60 days at one Munich site.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One site; short recording scheme; performance depends on settings.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0662-reported-measure","evaluation_data":null},{"case_id":"AIBD-0544","case_url":"records/AIBD-0544.html","measure":null,"reported_result":null,"baseline":"Advice available by phone in office hours or in person","result":"Advice available at any hour on WhatsApp and the web in more than 40 languages","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.bradford.gov.uk/adult-social-care/living-independently/annie-ai-digital-advisor/","https://www.local.gov.uk/case-studies/bradford-council-supporting-asc-front-door-ai-digital-assistants"],"source_refs":["AIBD-0544-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0544-reported-measure","evaluation_data":null},{"case_id":"AIBD-0553","case_url":"records/AIBD-0553.html","measure":null,"reported_result":null,"baseline":null,"result":"40 per cent+ increase in householder and minor applications processed each month per officer","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://web.archive.org/web/20261017035106/https://www.local.gov.uk/case-studies/leeds-city-council-and-xylo-transforming-planning-ai","https://www.local.gov.uk/case-studies/leeds-city-council-and-xylo-transforming-planning-ai","https://www.linkedin.com/posts/habin-isa_our-pilot-officers-processed-40-more-planning-activity-7437816039079489537-z4dP","https://www.gov.uk/algorithmic-transparency-records/leeds-city-council-xylo-core","https://www.linkedin.com/posts/buildxylo_using-xylo-cut-down-the-time-i-spent-on-activity-7437461682311925760-Oc3w","https://web.archive.org/web/20260423081109/https://www.ukauthority.com/articles/cambridge-leeds-and-plymouth-uses-of-proptech-shortlisted"],"source_refs":["AIBD-0553-source-0001","AIBD-0553-source-0002","AIBD-0553-source-0003","AIBD-0553-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0553-reported-measure","evaluation_data":null},{"case_id":"AIBD-0554","case_url":"records/AIBD-0554.html","measure":"Average days from receipt to validation and from validation to decision.","reported_result":"fell from 15.8 to 7.6 days","baseline":"15.8 days from receipt to validation and 53.1 days from validation to decision","result":"7.6 days from receipt to validation and 43.2 days from validation to decision over the pilot period","baseline_value":null,"result_value":null,"unit":null,"effect":"Receipt to validation fell from 15.8 to 7.6 days; validation to decision fell from 53.1 to 43.2 days.","design":"before/after.","sample":"Milton Keynes City Council planning service, October to December 2025; the number of applications is not stated.","uncertainty":"The council publishes no sample sizes and no rules for selecting the applications. The two accounts give different validation figures (7.6 and 7.7 days).","comparison_basis":"measured","source_urls":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/milton-keynes-city-council-and-valon-streamlining-planning-ai","https://www.local.gov.uk/case-studies/milton-keynes-city-council-and-valon-streamlining-planning-ai","https://www.valon-ai.com/case-studies/milton-keynes"],"source_refs":["https://web.archive.org/web/2026/https://www.local.gov.uk/case-studies/milton-keynes-city-council-and-valon-streamlining-planning-ai"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated. The council reports no control for case mix or staffing.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"The period before the pilot; Valon also cites a non-pilot group for decision times.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Receipt to validation fell from 15.8 to 7.6 days; validation to decision fell from 53.1 to 43.2 days.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"A nominated pilot group of validation and development management officers.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"The council analyzed its public planning register and officer timesheets.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Average days from receipt to validation and from validation to decision.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The planning register is public. The council's timesheets and analysis are not published.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Milton Keynes City Council planning service, October to December 2025; the number of applications is not stated.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The council publishes no sample sizes and no rules for selecting the applications. The two accounts give different validation figures (7.6 and 7.7 days).","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0554-reported-measure","evaluation_data":null},{"case_id":"AIBD-0650","case_url":"records/AIBD-0650.html","measure":"Time to write up meeting notes; staff hours saved.","reported_result":"an estimated 266,667 hours saved","baseline":"Usual write-up time (not stated in minutes)","result":"About 5 minutes of review and editing; a 73% average reduction","baseline_value":null,"result_value":null,"unit":null,"effect":"73% average reduction reported by pilot staff; about 266,667 hours saved (HMPPS estimate).","design":"before/after.","sample":"Pilot of 40 officers (30 survey responses), May to July 2025; over 1,600,000 meetings from 7 October 2025 to 14 September 2026.","uncertainty":"Self-reported pilot data; national figures rest on modeling assumptions; inaccuracies found in 20% of final case records examined.","comparison_basis":"measured","source_urls":["https://www.nao.org.uk/wp-content/uploads/2026/10/Productivity-improvement-the-Justice-Transcribe-AIenabled-transcription-tool.pdf","https://www.nao.org.uk/insights/productivity-improvement-the-justice-transcribe-ai-enabled-transcription-tool/","https://insidetime.org/newsround/probation-service-praised-for-positive-use-of-ai-in-meetings-with-clients/"],"source_refs":["AIBD-0650-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Officers' usual write-up time before the tool.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"73% average reduction reported by pilot staff; about 266,667 hours saved (HMPPS estimate).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Officers used the tool in supervision meetings.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"before/after.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"End-of-pilot survey of officers; national estimates from usage counts, with no data on actual time saved.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Time to write up meeting notes; staff hours saved.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"None public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Pilot of 40 officers (30 survey responses), May to July 2025; over 1,600,000 meetings from 7 October 2025 to 14 September 2026.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Self-reported pilot data; national figures rest on modeling assumptions; inaccuracies found in 20% of final case records examined.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0650-reported-measure","evaluation_data":null},{"case_id":"AIBD-0148","case_url":"records/AIBD-0148.html","measure":"over four and a half months the system had exposed three theft attempts","reported_result":"over four and a half months the system had exposed three theft attempts","baseline":null,"result":"nem nőtt meg a leltárhiány","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://profitline.hu/ujabb-hibrid-okosboltok-kialakitasat-tervezi-a-coop-szeged-zrt-479176","https://index.hu/gazdasag/2025/10/09/magyar-okosbolt-szeged-ai-lopas-lebuktatas/","https://www.minuszos.hu/ai-leplezett-le-tolvajokat-a-szegedi-hibrid-okosboltban/","https://turizmusonline.hu/cikkek/gasztro/a-szegediek-szeretik-az-okosboltokat--ujabbakat-nyitnak.html","https://www.economx.hu/vallalatok/2026/09/30/coop-szupermarket-szeged-szemelyzet-nelkuli-laurel/","https://24.hu/belfold/2025/10/08/okosbolt-szeged-coop-biztonsagi-or-lopas/","https://index.hu/gazdasag/2025/10/08/szeged-szemelyzet-nelkuli-bolt-lopas-biztonsagi-or/"],"source_refs":["AIBD-0148-source-0001","AIBD-0148-source-0002","AIBD-0148-source-0003","AIBD-0148-source-0004","AIBD-0148-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"inventory shrinkage under the store's previous staffed-only opening hours","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Result: nem nőtt meg a leltárhiány.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"over four and a half months the system had exposed three theft attempts","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0148-reported-measure","evaluation_data":null},{"case_id":"AIBD-0149","case_url":"records/AIBD-0149.html","measure":"found a missing 16-year-old in crisis within a day","reported_result":"found a missing 16-year-old in crisis within a day","baseline":null,"result":"Erfolg gab es nur einmal, als eine vermisste 16-Jährige mit der Kamera-KI identifiziert wurde.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://netzpolitik.org/2026/rotlichtviertel-frankfurt-am-main-hier-analysiert-die-polizei-jedes-gesicht/","https://www.rheinmainverlag.de/2025/10/16/kuenstliche-intelligenz-lokalisiert-16-jaehrige-vermisste-im-frankfurter-bahnhofsviertel/","https://www.presseportal.de/blaulicht/pm/4970/6138810","https://www.faz.net/aktuell/rhein-main/frankfurt/frankfurter-polizei-findet-vermisstes-maedchen-mithilfe-von-ki-110735027.html","https://www.fnp.de/frankfurt/ki-technik-hilft-polizei-in-frankfurt-bei-suche-nach-vermisster-93990652.html","https://www.bild.de/regional/frankfurt/ki-rettet-vermisstes-maedchen-16-aus-drogen-sumpf-von-frankfurt-68f0b438672b2f7280fac3b4","https://www.hessenschau.de/gesellschaft/hessen-erfand-den-datenschutz--und-ist-heute-vorreiter-bei-videoueberwachung-mit-ki-v1,erfindung-datenschutz-hessen-100.html","https://freiheitsrechte.org/uploads/documents/Freiheit-im-digitalen-Zeitalter/Biometrische-Videoueberwachung-Frankfurt/2026-04-22_Klagebegruendung-an-das-Verwaltungsgericht-Frankfurt-am-Main.pdf","https://freiheitsrechte.org/en/themen/freiheit-im-digitalen-zeitalter/biometrische-videoueberwachung"],"source_refs":["AIBD-0149-source-0001","AIBD-0149-source-0002","AIBD-0149-source-0003","AIBD-0149-source-0004","AIBD-0149-source-0005","AIBD-0149-source-0006","AIBD-0149-source-0007"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"all persons searched for with the Bahnhofsviertel face-recognition system since July 2025, a low double-digit number","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Result: Erfolg gab es nur einmal, als eine vermisste 16-Jährige mit der Kamera-KI identifiziert wurde.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"found a missing 16-year-old in crisis within a day","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0149-reported-measure","evaluation_data":null},{"case_id":"AIBD-0088","case_url":"records/AIBD-0088.html","measure":"Track forecast error, intensity forecast error, and forecast lead time.","reported_result":"track errors at 120 hours were less than half its five-year mean","baseline":"five-year mean official track errors of 58.7 n mi at 48 h, 128.7 n mi at 96 h and 185.4 n mi at 120 h","result":"Melissa official track errors of 40.3 n mi at 48 h, 66.2 n mi at 96 h and 87.7 n mi at 120 h; the GDMI ensemble mean alone had 32.6, 63.4 and 113.2 n mi","baseline_value":null,"result_value":null,"unit":null,"effect":"GDMI’s track forecasts were comparable to NHC’s official forecasts through 96 hours. NHC performed slightly better at 120 hours and had lower intensity errors than every guidance model through 96 hours.","design":"Retrospective verification of operational forecasts against the storm’s observed track and intensity.","sample":"Hurricane Melissa in the Atlantic, 21–31 October 2025. The report compares forecasts at lead times through 120 hours.","uncertainty":"The sources establish that forecasters used the guidance. They do not establish that it changed protective actions, reduced losses, or prevented deaths.","comparison_basis":"measured","source_urls":["https://www.nhc.noaa.gov/data/tcr/AL132025_Melissa.pdf","https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.001.shtml","https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.002.shtml","https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.008.shtml","https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.010.shtml","https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.013.shtml","https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.014.shtml","https://www.nhc.noaa.gov/archive/2025/al13/al132025.update.10281700.shtml","https://www.weather.gov/news/261102-AI-Hurricane-Forecasting","https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf","https://www.cnn.com/2025/12/03/climate/hurricane-forecast-melissa-ai-prediction-rapid-intensification","https://michaelrlowry.substack.com/p/this-hurricane-season-two-forecast"],"source_refs":["AIBD-0088-source-0010","AIBD-0088-source-0008","AIBD-0088-source-0009"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"NHC credits GDMI and regional models jointly with increasing forecasters’ confidence at unusually long lead times. It does not isolate GDMI’s effect on the official forecast.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"NHC’s official forecasts, conventional model guidance, and climatology-persistence forecasts for the same storm.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective verification of operational forecasts against the storm’s observed track and intensity.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"GDMI’s track forecasts were comparable to NHC’s official forecasts through 96 hours. NHC performed slightly better at 120 hours and had lower intensity errors than every guidance model through 96 hours.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The profile identifies an existing evaluations record, but exposure denominator has not been entered with a supported value.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"NHC’s forecast verification for Hurricane Melissa (AL132025).","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"NHC compared forecasts at matching lead times with its final estimates of the storm’s position and intensity. It compared official forecasts with model guidance on a common sample.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Track forecast error, intensity forecast error, and forecast lead time.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The profile identifies an existing evaluations record, but reproduction materials has not been entered with a supported value.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Hurricane Melissa in the Atlantic, 21–31 October 2025. The report compares forecasts at lead times through 120 hours.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"The sources establish that forecasters used the guidance. They do not establish that it changed protective actions, reduced losses, or prevented deaths.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0088-reported-measure","evaluation_data":null},{"case_id":"AIBD-0163","case_url":"records/AIBD-0163.html","measure":"at least £30m a year in avoided imbalance costs","reported_result":"at least £30m a year in avoided imbalance costs","baseline":"NESO's previous solar forecasts, with error levels not published","result":"large errors halved and accuracy 2.8 times higher, by the maker's account","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.openclimatefix.org/insights/neso-adopts-ai-solar-forecasting-control-room","https://theenergyst.com/national-energy-system-operator-adopts-ai-solar-forecasting-to-save-30m/","https://www.pv-magazine.com/2025/11/07/ai-powered-solar-forecasting-helps-uk-grid-operator-reduce-balancing-costs/","https://openclimatefix.org/","https://www.openclimatefix.org/insights/neso","https://www.iea.org/reports/modernising-grids-in-the-age-of-electricity/ai-enhanced-solutions"],"source_refs":["AIBD-0163-source-0002","AIBD-0163-source-0003","AIBD-0163-source-0004","AIBD-0163-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"NESO's solar forecasts before Quartz Solar was adopted","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: forecasting errors before adoption (not quantified). Result: halved forecasting errors.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"at least £30m a year in avoided imbalance costs","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0163-reported-measure","evaluation_data":null},{"case_id":"AIBD-0269","case_url":"records/AIBD-0269.html","measure":"an ejection fraction of 40% and severe aortic stenosis","reported_result":"an ejection fraction of 40% and severe aortic stenosis","baseline":"the accuracy of the cardiologists was 64.0% (95% CI 61.6–66.4%) with sensitivity 61.1% (95% CI 57.3–64.8%) and specificity 66.1% (95% CI 63.0–69.1%)","result":"an accuracy of 77.3%, sensitivity 72.6% and specificity of 80.7%","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12328201/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13080615/","https://www.nyp.org/news/ai-tool-developed-at-newyork-presbyterian-and-columbia-to-detect-hidden-heart-disease-receives-fda-approval","https://pmc.ncbi.nlm.nih.gov/articles/PMC13617667/","https://www.accessdata.fda.gov/cdrh_docs/pdf26/K260332.pdf"],"source_refs":["AIBD-0269-source-0001","AIBD-0269-source-0002","AIBD-0269-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"cardiologists reading the same ECGs without AI assistance (1,600 reviews of 150 ECGs)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: the accuracy of the cardiologists was 64.0% (95% CI 61.6–66.4%) with sensitivity 61.1% (95% CI 57.3–64.8%) and specificity 66.1% (95% CI 63.0–69.1%). Result: an accuracy of 77.3%, sensitivity 72.6% and specificity of 80.7%.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"echo showed LVEF 40% and severe aortic stenosis","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0269-reported-measure","evaluation_data":null},{"case_id":"AIBD-0342","case_url":"records/AIBD-0342.html","measure":"estimated emissions of 1,200 tonnes","reported_result":"estimated emissions of 1,200 tonnes","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://wedocs.unep.org/bitstreams/558dda69-c8e5-42c2-b10a-7b9c3dcc3e79/download","https://www.thechemicalengineer.com/news/un-satellite-programme-curbs-more-than-40-methane-super-emitters-since-2023/","https://methanedata.unep.org/downloads/top50_emitters/latest/mars_snapshot_latest.pdf","https://www.unep.org/topics/energy/methane/methane-alert-and-response-system-mars","https://web.archive.org/web/20261022235159/https://www.unep.org/en/what-we-do/energy/methane/mars-case-studies"],"source_refs":["AIBD-0342-source-0001","AIBD-0342-source-0002","AIBD-0342-source-0003","AIBD-0342-source-0004","AIBD-0342-source-0005","AIBD-0342-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"estimated emissions of 1,200 tonnes","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"estimated emissions of 1,200 tonnes","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0342-reported-measure","evaluation_data":null},{"case_id":"AIBD-0158","case_url":"records/AIBD-0158.html","measure":"halved Rosie's largest tumor by early 2026","reported_result":"halved Rosie's largest tumor by early 2026","baseline":"slowed but failed to shrink the tumours","result":"the tennis ball-sized tumour on Rosie’s hock has shrunk in half","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://stockhead.com.au/tech/tech-boss-uses-ai-and-chatgpt-to-create-cancer-vaccine-for-his-dying-dog/","https://news.unsw.edu.au/en/meet-the-man-who-designed-a-cancer-vaccine-for-his-dog","https://fortune.com/2026/03/15/australian-tech-entrepreneur-ai-cancer-vaccine-dog-rosie-unsw-mrna/","https://theconversation.com/a-man-used-ai-to-help-make-a-cancer-vaccine-for-his-dog-an-oncologist-urges-caution-278735","https://www.the-scientist.com/chatgpt-and-alphafold-help-design-personalized-vaccine-for-dog-with-cancer-74227","https://phys.org/news/2026-03-dog-chatgpt-australia-ai-vaccine.html","https://www.newsweek.com/owner-no-medical-background-invents-cure-dogs-terminal-cancer-11684882","https://reason.com/2026/03/19/man-successfully-designs-mrna-vaccine-to-treat-his-dogs-cancer/","https://www.abc.net.au/news/2026-06-22/australian-dog-cancer-treated-via-mrna-vaccine-and-ai/106692668","https://www.medicaldaily.com/rosie-dog-ai-mrna-cancer-vaccine-gamgee-launch-477500","https://finance.yahoo.com/healthcare/articles/one-dogs-legacy-becomes-gamgee-190000768.html","https://www.theverge.com/ai-artificial-intelligence/896878/ai-did-not-cure-this-dogs-cancer","https://www.36kr.com/p/3952021490269573"],"source_refs":["AIBD-0158-source-0001","AIBD-0158-source-0002","AIBD-0158-source-0003","AIBD-0158-source-0004","AIBD-0158-source-0005","AIBD-0158-source-0006","AIBD-0158-source-0007","AIBD-0158-source-0008","AIBD-0158-source-0009","AIBD-0158-source-0010","AIBD-0158-source-0011"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: veterinary chemotherapy and surgery given before the vaccine","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: slowed but failed to shrink the tumours. Result: the tennis ball-sized tumour on Rosie’s hock has shrunk in half.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"halved Rosie's largest tumor by early 2026","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0158-reported-measure","evaluation_data":null},{"case_id":"AIBD-0137","case_url":"records/AIBD-0137.html","measure":null,"reported_result":null,"baseline":"examined, prescribed acid blockers and sent home","result":"a CT the next day showed an appendix close to rupturing, which was removed","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.teslarati.com/man-credits-grok-ai-with-saving-his-life-after-er-missed-near-ruptured-appendix/","https://www.reddit.com/r/grok/comments/1pdzrk1/how_a_latenight_conversation_with_grok_got_me_to/","https://x.com/i/grok/share/RNHdX8qbyao4I4oNl6TEcmuNr","https://x.com/mark_k/status/1996857239311634793","https://www.latestly.com/socially/social-viral/man-says-xais-grok-helped-save-his-life-after-er-missed-near-ruptured-appendix-7229363.html","https://itavisen.no/2025/12/07/nordmann-grok-reddet-livet-blindtarm/","https://timesofindia.indiatimes.com/etimes/trending/49-year-old-man-says-grok-ai-saved-his-life-after-doctors-missed-his-appendix-emergency/articleshow/125819004.cms","https://www.tipranks.com/news/man-says-that-grok-ai-saved-his-life-after-an-er-misdiagnosis","https://x.com/elonmusk/status/2006834169637384470","https://www.ibtimes.co.uk/how-grok-saved-mans-life-when-doctors-couldnt-elon-musk-reacts-1767583","https://pmc.ncbi.nlm.nih.gov/articles/PMC12953526/"],"source_refs":["AIBD-0137-source-0001","AIBD-0137-source-0002","AIBD-0137-source-0003","AIBD-0137-source-0004","AIBD-0137-source-0005","AIBD-0137-source-0006","AIBD-0137-source-0007","AIBD-0137-source-0008","AIBD-0137-source-0009","AIBD-0137-source-0010"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0137-reported-measure","evaluation_data":null},{"case_id":"AIBD-0180","case_url":"records/AIBD-0180.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.medianama.com/2026/02/223-i4c-portal-banks-onboarded-amit-shah-mulehunter-ai/","https://www.medianama.com/2025/12/223-rti-23-banks-mulehunter-mule-accounts/","https://www.pib.gov.in/PressReleasePage.aspx?PRID=2260277","https://docs.rbihub.in/mule-hunter","https://economictimes.indiatimes.com/industry/banking/finance/banking/mule-hunter-tool-to-check-digital-fraud-is-showing-good-success-rate-rbi-governor-sanjay-malhotra/articleshow/125465527.cms","https://www.medianama.com/2026/08/223-sc-rbi-sop-review-compensation-digital-arrests/"],"source_refs":["AIBD-0180-source-0001","AIBD-0180-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0180-reported-measure","evaluation_data":null},{"case_id":"AIBD-0378","case_url":"records/AIBD-0378.html","measure":"Agreement score from 0 to 5 per phase and percent agreement per metric.","reported_result":"stent diameter (90.3% vs. 72.2%, P<0.05) and length selection (80.6% vs. 52.8%, P<0.01)","baseline":"72.2% agreement on stent diameter and 52.8% on stent length for junior physicians","result":"90.3% agreement on stent diameter and 80.6% on stent length for CA-GPT","baseline_value":null,"result_value":null,"unit":null,"effect":"Median pre-PCI agreement of 5 for CA-GPT against 3 for ChatGPT-5 and 4 for junior physicians; stent diameter agreement 90.3% against 72.2% for junior physicians.","design":"controlled observational.","sample":"96 patients with 160 lesions, from 156 consecutive patients who underwent OCT-guided PCI at Tangdu Hospital, Xi'an, between June 2024 and August 2025.","uncertainty":"Single center, retrospective, no clinical outcomes, a reference standard drawn from the same hospital's senior operators, different inputs for CA-GPT and ChatGPT-5, and authors who built the system.","comparison_basis":"measured","source_urls":["https://arxiv.org/abs/2512.10702","https://k.sina.com.cn/article_3996876140_ee3b7d6c0010166ey.html?from=tech","http://zj.xinhuanet.com/20251213/4d2d4be684c642d3b949c2666cae14c9/c.html","https://oecd.ai/en/incidents/2025-12-13-c088","https://github.com/deepseek-ai/DeepSeek-R1","https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B"],"source_refs":["https://arxiv.org/pdf/2512.10702"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"CA-GPT against junior physicians: 18.1 percentage points on stent diameter and 27.8 points on stent length; the authors did not test physicians assisted by the system.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"ChatGPT-5 given manually extracted OCT measurements, and junior interventional cardiologists who reviewed the full OCT images.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Median pre-PCI agreement of 5 for CA-GPT against 3 for ChatGPT-5 and 4 for junior physicians; stent diameter agreement 90.3% against 72.2% for junior physicians.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"None; the decisions were generated after the procedures and not acted on.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Decisions from CA-GPT, ChatGPT-5 and junior physicians scored for agreement with the procedural record of senior operators on ten pre-specified metrics, five before and five after PCI.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Agreement score from 0 to 5 per phase and percent agreement per metric.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"The authors state that all data are in the article and its supplementary files; the preprint offers no code or model weights for CA-GPT.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"96 patients with 160 lesions, from 156 consecutive patients who underwent OCT-guided PCI at Tangdu Hospital, Xi'an, between June 2024 and August 2025.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Single center, retrospective, no clinical outcomes, a reference standard drawn from the same hospital's senior operators, different inputs for CA-GPT and ChatGPT-5, and authors who built the system.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0378-reported-measure","evaluation_data":null},{"case_id":"AIBD-0152","case_url":"records/AIBD-0152.html","measure":null,"reported_result":null,"baseline":"The family searched without result and reported him missing at 8:15 p.m.","result":"The camera trace gave police his route and pictures for a public appeal, and a passer-by found him at about 3 a.m. and took him home","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://vnexpress.net/camera-ai-ho-tro-tim-cu-ong-di-lac-o-ha-noi-4994889.html","https://vietnamnet.vn/canh-sat-trich-xuat-camera-ai-giup-truy-vet-tim-nguoi-dan-ong-di-lac-o-ha-noi-2473262.html","https://tuoitre.vn/camera-ai-truy-vet-tim-cu-ong-di-lac-qua-nhieu-tuyen-pho-20251217102042429.htm","https://tienphong.vn/camera-ai-theo-dau-cu-ong-di-lac-o-ha-noi-post1805284.tpo","https://vov.vn/xa-hoi/camera-ai-tim-nguoi-di-lac-o-ha-noi-post1254446.vov","https://www.phunuonline.com.vn/tu-hinh-anh-qua-camera-ai-canh-sat-giup-cu-ong-di-lac-ve-nha-an-toan-a1569649.html","https://suckhoedoisong.vn/truy-vet-xuyen-dem-bang-camera-ai-kip-thoi-cuu-cu-ong-di-lac-bi-ha-duong-huyet-169251217065617017.htm"],"source_refs":["AIBD-0152-source-0002","AIBD-0152-source-0001","AIBD-0152-source-0003","AIBD-0152-source-0004","AIBD-0152-source-0005","AIBD-0152-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0152-reported-measure","evaluation_data":null},{"case_id":"AIBD-0153","case_url":"records/AIBD-0153.html","measure":"two pilots unhurt","reported_result":"two pilots unhurt","baseline":"a rapid loss of pressurization at about 23,000 ft, with both pilots on oxygen","result":"an automatic descent and landing at Rocky Mountain Metropolitan Airport with no injuries","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://avweb.com/aviation-news/garmin-autoland-activation-crew-decision/","https://www.cbsnews.com/colorado/news/plane-emergency-landing-colorado-autoland/","https://abcnews.com/US/small-plane-lands-safely-autoland-system-after-pilot/story?id=128629603","https://www.theregister.com/2025/12/24/autocomms/","https://www.denver7.com/news/front-range/broomfield/in-a-first-plane-makes-emergency-landing-at-rmma-using-automated-technology-due-to-pilot-incapacitation","https://avweb.com/aviation-news/garmin-autoland-safely-lands-king-air/","https://www.yahoo.com/news/articles/airplane-automatically-lands-itself-inflight-044612392.html","https://avbrief.com/new-recordings-include-king-air-autoland-pilot-transmissions/"],"source_refs":["AIBD-0153-source-0001","AIBD-0153-source-0002","AIBD-0153-source-0003","AIBD-0153-source-0004","AIBD-0153-source-0005","AIBD-0153-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"two pilots unhurt","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"two pilots unhurt","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0153-reported-measure","evaluation_data":null},{"case_id":"AIBD-0154","case_url":"records/AIBD-0154.html","measure":null,"reported_result":null,"baseline":"The footage showed only that the thief was a young man wearing a mask","result":"Further police work with AI face analysis built a profile of Phạm Ngọc Huy, and he was detained within about two days","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://dantri.com.vn/phap-luat/canh-sat-dung-ai-phan-tich-khuon-mat-bat-ke-trom-o-tphcm-20251225105544297.htm","https://cand.com.vn/Ban-tin-113/tu-du-lieu-cua-camera-ai-bat-doi-tuong-trom-sau-72-gio-gay-an-i792189/","https://vnexpress.net/camera-ai-giup-canh-sat-nhan-dang-ke-dot-nhap-phong-tro-4998114.html","https://tuoitre.vn/ung-dung-camera-ai-cong-an-phuong-tan-hung-bat-nhanh-ke-trom-deo-khau-trang-20251225112926111.htm","https://thanhnien.vn/cong-an-tphcm-lam-ro-vu-trom-cap-chi-sau-72-gio-nho-camera-ai-185251225111822364.htm","https://vietnamnet.vn/bat-nhanh-ke-trom-deo-khau-trang-nho-camera-ai-duong-pho-o-tphcm-2475850.html"],"source_refs":["AIBD-0154-source-0002","AIBD-0154-source-0003","AIBD-0154-source-0004","AIBD-0154-source-0005","AIBD-0154-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0154-reported-measure","evaluation_data":null},{"case_id":"AIBD-0023","case_url":"records/AIBD-0023.html","measure":"began cardiopulmonary resuscitation about 6–7 minutes after the collapse","reported_result":"began cardiopulmonary resuscitation about 6–7 minutes after the collapse","baseline":"Zonder die melding had ik hem niet zo snel gevonden (without that alert I would not have found him so quickly)","result":"CPR approximately 6–7 min after collapse; full neurological recovery","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.menshealth.nl/nieuws/hartstilstand-tijdens-hardlopen-apple-watch-redt-leven","https://europepmc.org/article/PMC/PMC13543876","https://www.want.nl/apple-watch-leven-gered-alwi-siebers/","https://112-meldingen.nl/112-varsseveld/incident-oude-doetinchemseweg-varsseveld-1402058","https://support.apple.com/en-us/108896"],"source_refs":["AIBD-0023-source-0001","AIBD-0023-source-0002","AIBD-0023-source-0003","AIBD-0023-source-0004","AIBD-0023-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: being found without the automatic alert, on a remote spot behind a hedge","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: Zonder die melding had ik hem niet zo snel gevonden (without that alert I would not have found him so quickly). Result: CPR approximately 6–7 min after collapse; full neurological recovery.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"CPR at 6–7 min, full neurological recovery","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0023-reported-measure","evaluation_data":null},{"case_id":"AIBD-0181","case_url":"records/AIBD-0181.html","measure":"raised assessed customs collections by over US$300 million from January to 17 July 2026","reported_result":"raised assessed customs collections by over US$300 million from January to 17 July 2026","baseline":"duties assessed on importers' declared values","result":"assessed collections 17.5 percent higher, over US$300 million from January to 17 July 2026, while the declared cost, insurance and freight value of imports rose about 6.3 percent","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://mofep.gov.gh/sites/default/files/budget-statements/2026-Mid-Year-Fiscal-Policy-Review-Speech.pdf","https://www.myjoyonline.com/gra-launches-ai-powered-system-to-boost-revenue-and-curb-customs-fraud/","https://oecd.ai/en/incidents/2026-01-16-e77d","https://ghanaiantimes.com.gh/ai-trade-analytics-rollout-to-boost-customs-revenue-by-45-gra-commissioner-general/","https://www.myjoyonline.com/publican-ai-boosts-gra-customs-revenue-to-gh%C2%A26-1bn-in-july-commissioner-general/","https://www.myjoyonline.com/guta-raises-alarm-over-alleged-sharp-increase-in-port-duties-under-publican-ai-system/","https://www.myjoyonline.com/okyere-baafi-calls-for-suspension-of-publican-ai-system-over-serious-flaws/","https://www.myjoyonline.com/publican-ai-only-provides-suggestions-not-final-decisions-gra-clarifies/"],"source_refs":["AIBD-0181-source-0002","AIBD-0181-source-0003","AIBD-0181-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: manual assessment and classification of goods by customs officers","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: could take up to two hours after importers submitted their documentation. Result: assessments could be completed within five minutes.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the system has already recovered GH¢15 million from fraudulent importers","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0181-reported-measure","evaluation_data":null},{"case_id":"AIBD-0255","case_url":"records/AIBD-0255.html","measure":"an overnight repair with no delay to passengers and zero delay minutes","reported_result":"an overnight repair with no delay to passengers and zero delay minutes","baseline":"over 10,000 delay minutes, multiple cancellations, and led to a full day of disruption for customers","result":"an overnight repair, with no delay to passengers, and zero delay minutes caused","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.newcivilengineer.com/latest/ai-system-on-lner-train-misidentifies-snake-as-track-defect-06-07-2026/","https://www.networkrail.co.uk/stories/future-rail-how-machine-learning-is-improving-your-journeys/","https://news.lner.co.uk/news/keeping-customers-moving-the-technology-helping-lner-and-network-rail-run-an-on-time-railway"],"source_refs":["AIBD-0255-source-0001","AIBD-0255-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a track defect reported by a train driver a week earlier (January 2026, Cambridgeshire)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: over 10,000 delay minutes, multiple cancellations, and led to a full day of disruption for customers. Result: an overnight repair, with no delay to passengers, and zero delay minutes caused.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"an overnight repair with no delay to passengers and zero delay minutes","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0255-reported-measure","evaluation_data":null},{"case_id":"AIBD-0309","case_url":"records/AIBD-0309.html","measure":null,"reported_result":null,"baseline":null,"result":"incidenten sneller te zien en te voorkomen","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.spoorpro.nl/actueel/2026/02/02/prorail-zet-288-nog-slimmere-cameras-in-langs-het-spoor/","https://www.spoorpro.nl/innovatie/2026/03/03/prorail-centraliseert-7-000-cameras-in-een-uniform-platform-voor-spoortoezicht/","https://www.prorail.nl/nieuws/prorail-zet-innovatief-cameratoezicht-in-voor-een-veilig-en-betrouwbaar-spoor","https://www.genetec.com/customer-stories/prorail"],"source_refs":["AIBD-0309-source-0002","AIBD-0309-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: ProRail's previous lineside camera surveillance (ProRail's own assertion)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Result: incidenten sneller te zien en te voorkomen.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0309-reported-measure","evaluation_data":null},{"case_id":"AIBD-0185","case_url":"records/AIBD-0185.html","measure":null,"reported_result":null,"baseline":"Rs 67.10 lakh transferred over several weeks, with more payments demanded to repay a loan before any withdrawal","result":"the victim recognized the fraud and went to the cybercrime police","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.newindianexpress.com/cities/bengaluru/2026/Jan/06/chatgpt-alert-helps-bengaluru-senior-citizen-uncover-rs-67-lakh-pre-ipo-investment-fraud","https://www.deccanherald.com/india/karnataka/bengaluru/bengaluru-chatgpt-response-saves-77-year-old-man-from-losing-more-money-to-cybercriminals-3853142"],"source_refs":["AIBD-0185-source-0001","AIBD-0185-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0185-reported-measure","evaluation_data":null},{"case_id":"AIBD-0585","case_url":"records/AIBD-0585.html","measure":null,"reported_result":null,"baseline":"Surveys with low response rates, research firms that cost thousands of dollars, and polite feedback from friends","result":"Instant answers from synthetic personas built from public data","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://blog.cambium.ai/introducing-chat-with-personas","https://blog.cambium.ai/market-validation-with-synthetic-personas","https://blog.cambium.ai/poll-your-ai-personas-by-audience-in-cambium-ai"],"source_refs":["AIBD-0585-source-0001","AIBD-0585-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0585-reported-measure","evaluation_data":null},{"case_id":"AIBD-0641","case_url":"records/AIBD-0641.html","measure":"Change in 6-minute walk distance; systolic blood pressure.","reported_result":"The app group gained 62.77 meters more in the 6-minute walk test than the control group","baseline":"Usual-care change in 6-minute walk distance","result":"62.77 m greater gain with the AI program","baseline_value":null,"result_value":null,"unit":null,"effect":"6MWD +62.77 m (95% CI 26.33-99.22); systolic BP 4.11 mmHg greater reduction (95% CI 0.11-8.28).","design":"randomized.","sample":"62 adults aged 18 to 75 with hypertension, Guangzhou.","uncertainty":"Small, open-label, convenience sample, eight weeks.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12848485"],"source_refs":["AIBD-0641-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Isolated against conventional guidance.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Conventional offline exercise guidance.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"6MWD +62.77 m (95% CI 26.33-99.22); systolic BP 4.11 mmHg greater reduction (95% CI 0.11-8.28).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"At least three sessions a week for eight weeks.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"6-minute walk test, cardiopulmonary exercise testing, blood pressure and questionnaires at baseline and 8 weeks.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Change in 6-minute walk distance; systolic blood pressure.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"CONSORT-eHEALTH checklist and supplementary tables.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"62 adults aged 18 to 75 with hypertension, Guangzhou.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Small, open-label, convenience sample, eight weeks.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0641-reported-measure","evaluation_data":null},{"case_id":"AIBD-0156","case_url":"records/AIBD-0156.html","measure":"heart rate climbed to 192 beats per minute","reported_result":"heart rate climbed to 192 beats per minute","baseline":"not stated for this patient; the reports give no time to detection without the monitor","result":"an alarm at the onset of ventricular tachycardia and CPR within the golden time","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.asiae.co.kr/article/2026021109383216218","https://it.chosun.com/news/articleView.html?idxno=2023092156866","https://www.wolyo.co.kr/news/articleView.html?idxno=308709","https://www.pinpointnews.co.kr/news/articleView.html?idxno=428348","https://www.newspim.com/news/view/20260211000281","https://view.asiae.co.kr/article/2026061909071730057","https://pmc.ncbi.nlm.nih.gov/articles/PMC10961211/"],"source_refs":["AIBD-0156-source-0002","AIBD-0156-source-0001","AIBD-0156-source-0003","AIBD-0156-source-0004","AIBD-0156-source-0005","AIBD-0156-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a heart rate that climbed to 192 beats per minute","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"a heart rate that climbed to 192 beats per minute","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0156-reported-measure","evaluation_data":null},{"case_id":"AIBD-0267","case_url":"records/AIBD-0267.html","measure":null,"reported_result":null,"baseline":"symmetric ossification defects adjacent to the sagittal suture that were not initially recognized by the treating physician","result":"suggested a diagnosis of \"Catlin mark skull,\" a historical term for Enlarged Parietal Foramina (EPF)","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=EXT_ID:41543485%20AND%20SRC:MED&resultType=core&format=json","https://pubmed.ncbi.nlm.nih.gov/41543485/","https://research.google/blog/advancing-medical-ai-with-med-gemini/"],"source_refs":["AIBD-0267-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the treating physician's initial reading of the CT scan","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: symmetric ossification defects adjacent to the sagittal suture that were not initially recognized by the treating physician. Result: suggested a diagnosis of \"Catlin mark skull,\" a historical term for Enlarged Parietal Foramina (EPF).","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0267-reported-measure","evaluation_data":null},{"case_id":"AIBD-0258","case_url":"records/AIBD-0258.html","measure":"called the fire service about 46 minutes before the operator reported","reported_result":"called the fire service about 46 minutes before the operator reported","baseline":"the operator's agent called police at about 11:50 on 20 January 2026","result":"the phone's automatic call reached the fire service at 11:04, about 46 minutes earlier","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.jiji.com/jc/article?k=2026012000646&g=soc","https://www.setn.com/News.aspx?NewsID=1784034","https://oecd.ai/en/incidents/2026-01-20-af0b","https://www.thestar.com.my/aseanplus/aseanplus-news/2026/02/18/three-bodies-found-at-helicopter-crash-site-in-southwest-japan-volcanic-crater","https://www.asahi.com/articles/ASV1N1TVWV1NTIPE00DM.html","https://japan.storm.mg/articles/1096781"],"source_refs":["AIBD-0258-source-0001","AIBD-0258-source-0002","AIBD-0258-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0258-reported-measure","evaluation_data":null},{"case_id":"AIBD-0298","case_url":"records/AIBD-0298.html","measure":"detected cardiac arrest at 09:50:17 after loss of pulse","reported_result":"detected cardiac arrest at 09:50:17 after loss of pulse","baseline":null,"result":"Cardiac arrest was detected at 09:50:17, after a loss of pulse in the photoplethysmography signal.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12914388/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13427576/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13182813/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12914390/","https://pmc.ncbi.nlm.nih.gov/articles/PMC13001639/"],"source_refs":["AIBD-0298-source-0001","AIBD-0298-source-0002","AIBD-0298-source-0003","AIBD-0298-source-0004","AIBD-0298-source-0005","AIBD-0298-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the ICD's electrograms and the confirmed time of collapse (velocity 0 km/h), reviewed in post-processing","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Result: Cardiac arrest was detected at 09:50:17, after a loss of pulse in the photoplethysmography signal.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"detected cardiac arrest at 09:50:17 after loss of pulse","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0298-reported-measure","evaluation_data":null},{"case_id":"AIBD-0299","case_url":"records/AIBD-0299.html","measure":null,"reported_result":null,"baseline":null,"result":"The algorithm accurately identified the moment of cardiac arrest in concordance with the clinical reference signals.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12914423/","https://doi.org/10.1016/j.resuscitation.2025.110657","https://pmc.ncbi.nlm.nih.gov/articles/PMC12914390/"],"source_refs":["AIBD-0299-source-0001","AIBD-0299-source-0002","AIBD-0299-source-0003","AIBD-0299-source-0004","AIBD-0299-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the clinical reference signals (ECG, invasive arterial pressure and clinical photoplethysmography) recorded alongside the watch","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Result: The algorithm accurately identified the moment of cardiac arrest in concordance with the clinical reference signals.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0299-reported-measure","evaluation_data":null},{"case_id":"AIBD-0155","case_url":"records/AIBD-0155.html","measure":"the fire burned ten square meters with nobody hurt","reported_result":"the fire burned ten square meters with nobody hurt","baseline":null,"result":"在火勢轉劇前一分鐘便偵測到煙霧","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://n.yam.com/Article/20260205197916","https://udn.com/news/story/7320/9310526","https://udn.com/news/story/7320/9310088","https://news.tvbs.com.tw/local/3120819","https://n.yam.com/Article/20260205444546","https://news.cts.com.tw/cts/general/202602/202602052559104.html","https://www.nownews.com/news/6783602","https://news.nextapple.com/local/20260205/E631BE3355F1C46E511BAEBE0A5D0D61","https://n.yam.com/Article/20260205590605","https://tw.news.yahoo.com/ai%E9%80%B2%E9%A7%90%E5%AE%AE%E5%BB%9F-%E7%B5%B1%E8%A8%88%E4%BA%BA%E6%B5%81-%E9%A0%90%E9%98%B2%E7%81%AB%E7%81%BD%E9%82%84%E8%83%BD-%E8%A7%A3%E7%B1%A4%E8%A9%A9-121500289.html","https://www.advantech.com/en-us/resources/case-study/%E7%AB%8B%E9%81%94%E8%BB%9F%E9%AB%94%E6%94%9C%E6%89%8B%E7%A0%94%E8%8F%AFai%E9%96%80%E7%A5%9E%EF%BC%9A%E5%89%B5%E6%96%B0%E6%8A%80%E8%A1%93%E5%8A%A9%E5%8A%9B%E5%AF%BA%E5%BB%9F%E6%96%87%E5%8C%96%E8%B3%87%E7%94%A2%E4%BF%9D%E8%AD%B7","https://udn.com/news/story/7326/9714008"],"source_refs":["AIBD-0155-source-0002","AIBD-0155-source-0001","AIBD-0155-source-0003","AIBD-0155-source-0004","AIBD-0155-source-0005","AIBD-0155-source-0006","AIBD-0155-source-0007","AIBD-0155-source-0008","AIBD-0155-source-0009","AIBD-0155-source-0010","AIBD-0155-source-0011","AIBD-0155-source-0012"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the temple's overnight duty staff relying on their own watch, without the AI smoke alert","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Result: 在火勢轉劇前一分鐘便偵測到煙霧.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"held to 10 m², no injuries, main hall intact","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0155-reported-measure","evaluation_data":null},{"case_id":"AIBD-0274","case_url":"records/AIBD-0274.html","measure":null,"reported_result":null,"baseline":"risk of classing the right-sided finding as isolated diaphragmatic elevation","result":"AI flag of bilateral effusion, with the right side confirmed as a subpulmonic effusion on ultrasound","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12970633/","https://pubmed.ncbi.nlm.nih.gov/37884394/"],"source_refs":["AIBD-0274-source-0001","AIBD-0274-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0274-reported-measure","evaluation_data":null},{"case_id":"AIBD-0160","case_url":"records/AIBD-0160.html","measure":"325 detained and remanded, 69 groups closed","reported_result":"325 detained and remanded, 69 groups closed","baseline":"insan gücüyle aylar sürecek bir analiz","result":"saniyeler içine sığdırılarak","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.aa.com.tr/tr/gundem/istanbul-emniyeti-yapay-zeka-destekli-avciyla-uyusturucu-saticilarinin-siber-inlerine-giriyor/3894439","https://www.haberler.com/guncel/istanbul-emniyeti-yapay-zeka-destekli-avci-yla-19720800-haberi/","https://www.aksam.com.tr/guncel/zehir-tacirleriyle-mucadelede-yeni-donem-avci-goz-actirmiyor/haber-1656417","https://www.hurriyet.com.tr/gundem/avci-uyusturucunun-siber-inlerinde-43145237","https://www.aa.com.tr/tr/gundem/istanbulda-uyusturucu-saticilarina-yonelik-operasyonda-yapay-zeka-avci-aktif-rol-aldi/4000945","https://www.aa.com.tr/tr/gundem/14-ilde-duzenlenen-uyusturucu-operasyonlarinda-305-supheli-yakalandi/3825356","https://www.aa.com.tr/tr/gundem/istanbul-merkezli-14-ildeki-uyusturucu-operasyonunda-324-supheli-tutuklandi/3829523"],"source_refs":["AIBD-0160-source-0001","AIBD-0160-source-0002","AIBD-0160-source-0003","AIBD-0160-source-0004","AIBD-0160-source-0005"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: manual analysis of messaging data by police staff","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: insan gücüyle aylar sürecek bir analiz. Result: saniyeler içine sığdırılarak.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"325 detained and remanded, 69 groups closed","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0160-reported-measure","evaluation_data":null},{"case_id":"AIBD-0102","case_url":"records/AIBD-0102.html","measure":"at least 39 poachers surrendered","reported_result":"at least 39 poachers surrendered","baseline":"Rangers did not know when poachers entered the reserve","result":"Rangers know the area where poachers are and can prepare a response","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.thehindu.com/sci-tech/energy-and-environment/ai-fit-cameras-in-similipal-tiger-reserve-send-poaching-plummeting/article69225924.ece","https://www.newindianexpress.com/states/odisha/2026/Feb/20/39-persons-arrested-as-ai-cameras-foil-major-poaching-bid-in-similipal","https://timesofindia.indiatimes.com/city/bhubaneswar/ai-camera-footage-helps-nab-5-armed-poachers-in-similipal/articleshow/130197981.cms","https://theprint.in/india/ai-enabled-camera-helps-forest-dept-to-nab-two-poachers-in-similipal-national-park/3052132/","https://theprint.in/india/ai-enabled-camera-helps-nab-four-poachers-in-similipal-tiger-reserve/2996083/","https://sambadenglish.com/latest-news/four-poachers-arrested-for-illegal-entry-into-similipal-wildlife-sanctuary-in-odishas-mayurbhanj-hunting-bid-foiled-12193080","https://www.thehansindia.com/news/national/ai-camera-helps-nab-four-poachers-in-similipal-1101741","https://ommcomnews.com/odisha-news/ai-trailguard-cameras-help-nab-three-poachers-in-similipal-wildlife-sanctuary/","https://www.newindianexpress.com/states/odisha/2026/Feb/27/half-of-similipal-tiger-reserve-to-come-under-ai-enabled-cameras"],"source_refs":["AIBD-0102-source-0001","AIBD-0102-source-0002","AIBD-0102-source-0003","AIBD-0102-source-0004","AIBD-0102-source-0005","AIBD-0102-source-0006","AIBD-0102-source-0007"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"at least 39 poachers surrendered","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"at least 39 poachers surrendered","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0102-reported-measure","evaluation_data":null},{"case_id":"AIBD-0157","case_url":"records/AIBD-0157.html","measure":null,"reported_result":null,"baseline":"would have likely crashed or caused a pileup","result":"immediately slowed, activated hazards, and safely pulled over to the shoulder","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.teslarati.com/tesla-cybertruck-fsd-saving-life-freeway-medical-emergency/","https://x.com/rishivohra/status/2024539159269167410","https://www.hindustantimes.com/trending/elon-musk-reacts-after-man-claims-tesla-saved-his-life-during-freeway-medical-emergency-life-saving-tech-101771574831376.html","https://www.storyboard18.com/trending/indian-origin-tesla-owner-says-car-prevented-crash-after-he-passed-out-musk-says-glad-youre-ok-90258.htm","https://americanbazaaronline.com/2026/02/20/elon-musk-responds-to-tesla-saving-a-persons-life-during-accident-475581/","https://x.com/elonmusk/status/2024551075576189357"],"source_refs":["AIBD-0157-source-0001","AIBD-0157-source-0002","AIBD-0157-source-0003","AIBD-0157-source-0004","AIBD-0157-source-0005","AIBD-0157-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: a manually driven vehicle whose driver loses consciousness","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: would have likely crashed or caused a pileup. Result: immediately slowed, activated hazards, and safely pulled over to the shoulder.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0157-reported-measure","evaluation_data":null},{"case_id":"AIBD-0302","case_url":"records/AIBD-0302.html","measure":null,"reported_result":null,"baseline":"Burns says that she did not think that anybody could hear her or knew what had happened.","result":"A dispatcher stayed on the line, and crews pulled her out with bruises only.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://abcnews.com/GMA/News/19-year-pulled-sinking-car-details-rescue-responders/story?id=130747244","https://www.wlwt.com/article/miami-river-body-camera-video-rescue-car-truck/70526120","https://web.archive.org/web/20260311210830/https://local12.com/news/local/woman-thanks-rescuers-who-saved-her-from-crash-into-clermont-county-creek-andi-burns-ice-little-miami-river-apple-watch-thomas-jetter-first-responders","https://support.apple.com/en-us/104959","https://pmc.ncbi.nlm.nih.gov/articles/PMC10363817/"],"source_refs":["AIBD-0302-source-0001","AIBD-0302-source-0002","AIBD-0302-source-0003","AIBD-0302-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0302-reported-measure","evaluation_data":null},{"case_id":"AIBD-0344","case_url":"records/AIBD-0344.html","measure":"an estimated 60,500 tonnes of methane emissions","reported_result":"an estimated 60,500 tonnes of methane emissions","baseline":"an estimated 1.7m tCO2e since 2023","result":"operators diverted gas production to a newer facility","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.thechemicalengineer.com/news/un-satellite-programme-curbs-more-than-40-methane-super-emitters-since-2023/","https://www.unep.org/topics/energy/methane/methane-alert-and-response-system-mars","https://wedocs.unep.org/bitstreams/5c3f0098-f3d6-402d-beca-83f2971e69b9/download"],"source_refs":["AIBD-0344-source-0001","AIBD-0344-source-0002","AIBD-0344-source-0003","AIBD-0344-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the flare before gas was diverted","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: an estimated 1.7m tCO2e since 2023. Result: operators diverted gas production to a newer facility.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"An estimated 60,500 tonnes of methane emitted","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0344-reported-measure","evaluation_data":null},{"case_id":"AIBD-0327","case_url":"records/AIBD-0327.html","measure":"one drone intercepted near Shlomi","reported_result":"one drone intercepted near Shlomi","baseline":"A single Iron Dome interceptor missile costs about US$50,000","result":"each shot cost around $US3.50, and more recent estimates suggest the cost may now have fallen as low as US$2.50 per shot","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["http://theconversation.com/israels-iron-beam-why-laser-weapons-are-no-longer-science-fiction-277390","https://www.timesofisrael.com/iron-dome-air-defenses-upgraded-incorporating-lessons-learned-from-iran-war/","https://oecd.ai/en/incidents/2026-02-28-d62f","https://www.ynetnews.com/article/h13axdmfwx","https://www.jpost.com/israel-news/defense-news/article-888515","https://www.armyrecognition.com/news/army-news/2026/israels-iron-beam-laser-linked-to-first-combat-intercepts-of-hezbollah-rockets-in-lebanon","https://www.timesnownews.com/world/middle-east/did-israel-debut-their-iron-beam-defence-system-to-intercept-missiles-fired-by-hezbollah-what-we-know-so-far-article-153731169","https://www.timesofisrael.com/defense-ministry-hands-idf-first-combat-ready-iron-beam-laser-interception-system/","https://www.timesofisrael.com/surge-in-drone-attacks-sharpens-focus-on-laser-defense/","https://www.timesofisrael.com/cheap-and-elusive-drones-put-incessant-pressure-on-israels-evolving-air-defenses/","https://en.wikipedia.org/wiki/Iron_Beam","https://english.hani.co.kr/arti/english_edition/e_international/1111552.html","https://www.jpost.com/defense-and-tech/article-889701","https://www.jpost.com/israel-news/defense-news/article-894838","https://www.ynetnews.com/article/s1uefqsifl"],"source_refs":["AIBD-0327-source-0002","AIBD-0327-source-0003","AIBD-0327-source-0004","AIBD-0327-source-0005","AIBD-0327-source-0006","AIBD-0327-source-0007","AIBD-0327-source-0008","AIBD-0327-source-0009","AIBD-0327-source-0010","AIBD-0327-source-0001","AIBD-0327-source-0011","AIBD-0327-source-0012"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: Iron Dome interceptor missiles","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: A single Iron Dome interceptor missile costs about US$50,000. Result: each shot cost around $US3.50, and more recent estimates suggest the cost may now have fallen as low as US$2.50 per shot.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"one drone intercepted near Shlomi","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0327-reported-measure","evaluation_data":null},{"case_id":"AIBD-0206","case_url":"records/AIBD-0206.html","measure":"went from idea to working exploit in under 48 hours","reported_result":"went from idea to working exploit in under 48 hours","baseline":"an impractical amount of time and manual effort (Wiz); months' worth of manual analysis, per The Register's account of Wiz","result":"from idea to working exploit in less than 48 hours","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.wiz.io/blog/github-rce-vulnerability-cve-2026-3854","https://github.blog/security/securing-the-git-push-pipeline-responding-to-a-critical-remote-code-execution-vulnerability/","https://www.cve.org/CVERecord?id=CVE-2026-3854","https://docs.github.com/en/enterprise-server@3.14/admin/release-notes","https://www.theregister.com/2026/04/29/github_woah_a_genuinely_helpful/","https://www.theverge.com/news/920295/github-remote-code-execution-vulnerability-fix","https://www.darkreading.com/application-security/reverse-engineering-ai-unearths-high-severity-github-bug","https://www.webpronews.com/wizs-ai-powered-strike-how-claude-cracked-githubs-git-fortress-for-a-fat-bounty/","https://code.claude.com/docs/en/overview","https://services.nvd.nist.gov/rest/json/cves/2.0?cveId=CVE-2026-3854"],"source_refs":["AIBD-0206-source-0002","AIBD-0206-source-0003","AIBD-0206-source-0004","AIBD-0206-source-0001","AIBD-0206-source-0005","AIBD-0206-source-0006","AIBD-0206-source-0007","AIBD-0206-source-0008","AIBD-0206-source-0009"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: manual reverse engineering of GitHub's compiled binaries without AI tooling","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: an impractical amount of time and manual effort (Wiz); months' worth of manual analysis, per The Register's account of Wiz. Result: from idea to working exploit in less than 48 hours.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"went from idea to working exploit in under 48 hours","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0206-reported-measure","evaluation_data":null},{"case_id":"AIBD-0204","case_url":"records/AIBD-0204.html","measure":"T18 reached 110 characters (22 words) per minute with a 1.6% word error rate","reported_result":"T18 reached 110 characters (22 words) per minute with a 1.6% word error rate","baseline":"the previous state-of-the-art method, 20 characters per minute slower (90 characters per minute, about 18 words per minute, per Scientific American); able-bodied smartphone typing of 27.3 words per minute for T18's age group","result":"110 characters per minute (22 words per minute) at a 1.6% word error rate","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.nature.com/articles/s41593-026-02218-y","https://europepmc.org/article/PMC/PMC13633126","https://neurosciencenews.com/brain-computer-interface-typing-paralysis-30325/","https://medicalxpress.com/news/2026-03-brain-interface-enables-rapid-communication.html","https://www.scientificamerican.com/article/brain-implant-allows-people-who-are-paralyzed-to-type-using-their-thoughts/","https://alsnewstoday.com/news/brain-implant-decodes-finger-moves-paralyzed-patients-type/","https://www.infosalus.com/salud-investigacion/noticia-neuroprotesis-experimental-permite-personas-paralisis-escribir-mente-si-usaran-teclado-20260317071649.html","https://clinicaltrials.gov/study/NCT00912041","https://pmc.ncbi.nlm.nih.gov/articles/PMC13335209/"],"source_refs":["AIBD-0204-source-0001","AIBD-0204-source-0002","AIBD-0204-source-0003","AIBD-0204-source-0004","AIBD-0204-source-0005","AIBD-0204-source-0006","AIBD-0204-source-0007","AIBD-0204-source-0008"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the previous state-of-the-art hand-motor iBCI (handwriting decoding) and able-bodied smartphone typing","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: the previous state-of-the-art method, 20 characters per minute slower (90 characters per minute, about 18 words per minute, per Scientific American); able-bodied smartphone typing of 27.3 words per minute for T18's age group. Result: 110 characters per minute (22 words per minute) at a 1.6% word error rate.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"T18 reached 110 characters (22 words) per minute with a 1.6% word error rate","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0204-reported-measure","evaluation_data":null},{"case_id":"AIBD-0584","case_url":"records/AIBD-0584.html","measure":"Mean absolute occupancy-probability differences were 0.04-0.05 in Washington, 0.05-0.07 in Montana, and 0.03-0.06 in Guatemala for species shared by the AI and expert datasets.","reported_result":"Mean occupancy differences of 0.03-0.07 across study areas","baseline":"Expert-based estimates of species occupancy and species-environment relationships.","result":"Mean absolute occupancy-probability differences of 0.04-0.05 in Washington, 0.05-0.07 in Montana, and 0.03-0.06 in Guatemala for shared species.","baseline_value":null,"result_value":null,"unit":null,"effect":"Mean occupancy estimates were broadly similar for shared species. Individual differences reached 0.40 in Montana, and some species were omitted or added by classification errors.","design":"Comparative ecological analysis of the same regional image datasets using expert labels and six AI processing configurations. The comparator is not uniformly free of AI.","sample":"Washington: 638 cameras and about 1.8 million images. Montana: 263 cameras and about 1.7 million images. Guatemala: 326 cameras and about 280,000 images.","uncertainty":"Comparisons of model estimates cover shared species. Rare-species omissions remain outside those comparisons. Regional filtering and sequence processing affect results. Montana's expert workflow already used AI. The exact SpeciesNet checkpoint, some regional inputs, and the complete supplement were unavailable in this review.","comparison_basis":"measured","source_urls":["https://doi.org/10.1111/1365-2664.70370","https://zenodo.org/records/19560799","https://doi.org/10.64898/2026.10.06.757089"],"source_refs":["AIBD-0584-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed"},"evaluations[].design":{"state":"value_present","review_state":"reviewed"},"evaluations[].effect":{"state":"value_present","review_state":"reviewed"},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed"},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed"},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Mean absolute occupancy-probability differences were 0.04-0.05 in Washington, 0.05-0.07 in Montana, and 0.03-0.06 in Guatemala for species shared by the AI and expert datasets.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Zenodo version 1.0.1 contains Washington CSV inputs and scripts for all regions. Referenced Montana and Guatemala inputs are absent. The public Washington image subset has a different denominator. The complete supplement was inaccessible.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Washington: 638 cameras and about 1.8 million images. Montana: 263 cameras and about 1.7 million images. Guatemala: 326 cameras and about 280,000 images.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed"}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0584-reported-measure","evaluation_data":{"mean_absolute_occupancy_probability_difference":{"Washington":[0.04,0.05],"Montana":[0.05,0.07],"Guatemala":[0.03,0.06]},"scope":"Species shared by expert and AI datasets, across six processing configurations.","maximum_individual_species_difference_Montana":0.4,"Montana_mountain_goat_detections_misclassified":94}},{"case_id":"AIBD-0653","case_url":"records/AIBD-0653.html","measure":"Adenoma detection rate.","reported_result":"Statistically significant improvement for most technologies; not significant for advanced adenomas or sessile serrated lesions.","baseline":"Adenoma detection rate with standard colonoscopy","result":"Higher pooled adenoma detection rate with AI for most systems","baseline_value":null,"result_value":null,"unit":null,"effect":"Statistically significant improvement for most technologies; not significant for advanced adenomas or sessile serrated lesions.","design":"randomized.","sample":"70 studies of AI-assisted live colonoscopy, mostly randomized controlled trials.","uncertainty":"Gains mostly in small non-advanced adenomas; effect on cancer rates uncertain; little data on false positives.","comparison_basis":"measured","source_urls":["https://www.nice.org.uk/guidance/htg773/chapter/3-Committee-discussion","https://www.nice.org.uk/guidance/htg773/chapter/1-Recommendations"],"source_refs":["AIBD-0653-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Higher adenoma detection rate overall.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Standard colonoscopy.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Statistically significant improvement for most technologies; not significant for advanced adenomas or sessile serrated lesions.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Colonoscopies with the software running.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Meta-analysis of the adenoma detection rate for each technology by the external assessment group.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Adenoma detection rate.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"External assessment group report on the NICE website.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"70 studies of AI-assisted live colonoscopy, mostly randomized controlled trials.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Gains mostly in small non-advanced adenomas; effect on cancer rates uncertain; little data on false positives.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0653-reported-measure","evaluation_data":null},{"case_id":"AIBD-0591","case_url":"records/AIBD-0591.html","measure":null,"reported_result":null,"baseline":"A consultant report costing $10,000 to $15,000 that takes weeks to produce","result":"A generated report in four minutes","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://thetabulareport.substack.com/p/seeing-the-whole-board","https://grantrail.org/"],"source_refs":["AIBD-0591-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0591-reported-measure","evaluation_data":null},{"case_id":"AIBD-0312","case_url":"records/AIBD-0312.html","measure":"contained it before it grew past 7 acres (2.8 hectares)","reported_result":"contained it before it grew past 7 acres (2.8 hectares)","baseline":"the first 911 call","result":"about 45 minutes faster on average than the first 911 call","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.courthousenews.com/states-across-the-wildfire-prone-western-us-are-using-ai-for-early-detection/","https://apnews.com/article/ai-wildfire-detection-cameras-194656fe63ea89dbc4661eaf8b79f6bb","https://azdailysun.com/news/local/seven-acre-fire-fully-contained-south-of-flagstaff-on-thursday-afternoon/article_585bebfe-c17c-4872-a37d-7e08195a39d7.html","https://services3.arcgis.com/T4QMspbfLg3qTGWY/arcgis/rest/services/WFIGS_Incident_Locations_YearToDate/FeatureServer/0/query?where=UniqueFireIdentifier%3D%272026-AZCOF-000289%27&outFields=IncidentName,UniqueFireIdentifier,POOJurisdictionalUnit,FireDiscoveryDateTime,DiscoveryAcres,IncidentSize,ContainmentDateTime,FireCause&f=pjson"],"source_refs":["AIBD-0312-source-0001","AIBD-0312-source-0002","AIBD-0312-source-0003","AIBD-0312-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the first 911 call reporting a fire","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: the first 911 call. Result: about 45 minutes faster on average than the first 911 call.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"contained it before it grew past 7 acres (2.8 hectares)","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0312-reported-measure","evaluation_data":null},{"case_id":"AIBD-0594","case_url":"records/AIBD-0594.html","measure":null,"reported_result":null,"baseline":"Not disclosed","result":"Not disclosed; the provider says only that the model predicted preferences more accurately","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.opendatalabs.com/case-studies/fortune-500-fashion-brand"],"source_refs":["AIBD-0594-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0594-reported-measure","evaluation_data":null},{"case_id":"AIBD-0159","case_url":"records/AIBD-0159.html","measure":"A police team reached his home within 12 minutes","reported_result":"A police team reached his home within 12 minutes","baseline":"escalation speed of other reports","result":"escalated to local authorities twice as quickly as other reports","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://about.fb.com/news/2017/11/getting-our-community-help-in-real-time/","https://ianslive.in/teens-life-saved-in-12-minutes-raebareli-police-act-swiftly-after-social-media-suicide-alert--20260401091324","https://timesofindia.indiatimes.com/city/lucknow/cops-save-18-yr-old-from-suicide-after-meta-alert/articleshow/129935943.cms","https://telanganatoday.com/teens-life-saved-in-12-minutes-up-police-act-swiftly-after-social-media-suicide-alert","https://english.mathrubhumi.com/news/good-news/up-police-rescue-teen-social-media-alert-ar7x14gi","https://www.decodeinternet.in/decode/meta-sent-suicide-alert-uttar-pradesh-police-had-minutes-to-save-30394","https://about.fb.com/news/2018/09/inside-feed-suicide-prevention-and-ai/","https://www.thehansindia.com/news/national/cops-save-man-in-8-mins-after-suicide-alert-on-insta-1084758"],"source_refs":["AIBD-0159-source-0002","AIBD-0159-source-0003","AIBD-0159-source-0001","AIBD-0159-source-0004","AIBD-0159-source-0005","AIBD-0159-source-0006","AIBD-0159-source-0007","AIBD-0159-source-0008"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"reports escalated without Meta's pattern-recognition acceleration","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: escalation speed of other reports. Result: escalated to local authorities twice as quickly as other reports.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A police team reached his home within 12 minutes","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0159-reported-measure","evaluation_data":null},{"case_id":"AIBD-0089","case_url":"records/AIBD-0089.html","measure":"beat its official Atlantic track forecasts at 12 to 72 hours","reported_result":"beat its official Atlantic track forecasts at 12 to 72 hours","baseline":"OFCL errors of 20.4, 31.1, 41.9, 49.2, 62.2 and 81.4 n mi at 12, 24, 36, 48, 60 and 72 h","result":"GDMI errors of 19.6, 26.4, 34.8, 41.4, 50.8 and 66.5 n mi at the same lead times","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf","https://www.weather.gov/news/261102-AI-Hurricane-Forecasting","https://www.cnn.com/2025/12/03/climate/hurricane-forecast-melissa-ai-prediction-rapid-intensification","https://michaelrlowry.substack.com/p/this-hurricane-season-two-forecast"],"source_refs":["AIBD-0089-source-0001","AIBD-0089-source-0002","AIBD-0089-source-0003","AIBD-0089-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the NHC official track forecast (OFCL) in the 2025 Atlantic homogeneous model comparison","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"NHC compared forecasts for the same eligible storms and lead times against its best-track analyses, official forecasts, and other guidance. Storms without GDMI forecasts early in the season were excluded from the relevant comparison.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: OFCL errors of 20.4, 31.1, 41.9, 49.2, 62.2 and 81.4 n mi at 12, 24, 36, 48, 60 and 72 h. Result: GDMI errors of 19.6, 26.4, 34.8, 41.4, 50.8 and 66.5 n mi at the same lead times.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"beat official Atlantic track forecasts at 12–72 h","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0089-reported-measure","evaluation_data":null},{"case_id":"AIBD-0205","case_url":"records/AIBD-0205.html","measure":null,"reported_result":null,"baseline":"turning up after the event to clean up and commiserate with devastated customers","result":"spotting hundreds of potential blockages before it is too late","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.dailyecho.co.uk/news/26045031.fatberg-found-gosport-removed-causing-damage/","https://www.andoveradvertiser.co.uk/news/26046397.southern-water-using-ai-prevent-pollution-incidents/","https://www.islandecho.co.uk/ai-sewer-tech-prevents-fatberg-from-polluting-chalk-stream-in-calbourne/","https://www.bbc.com/future/article/20260511-fatbergs-are-taking-over-city-sewers-can-we-stop-them"],"source_refs":["AIBD-0205-source-0001","AIBD-0205-source-0002","AIBD-0205-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: the previous reactive practice of attending blockages after a flooding or pollution event","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: turning up after the event to clean up and commiserate with devastated customers. Result: spotting hundreds of potential blockages before it is too late.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0205-reported-measure","evaluation_data":null},{"case_id":"AIBD-0328","case_url":"records/AIBD-0328.html","measure":"one rescued airman; the claimed AI contribution remains unverified","reported_result":"one rescued airman; the claimed AI contribution remains unverified","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://nypost.com/2026/04/07/us-news/ghost-murmur-a-never-used-secret-tool-deployed-to-find-lost-airman-in-iran-in-daring-mission/","https://www.govinfo.gov/content/pkg/DCPD-202600233/pdf/DCPD-202600233.pdf","https://nypost.com/2026/04/08/us-news/trump-confirms-cia-ghost-murmur-tool-was-very-important-to-find-airman-in-iran-as-experts-debate-how-it-works/","https://www.scientificamerican.com/article/what-is-the-quantum-ghost-murmur-purportedly-used-in-iran-scientists/","https://www.cbsnews.com/news/wounded-air-force-officer-evaded-capture-iran-60-minutes-transcript/","https://www.cbsnews.com/news/takeaways-air-force-officer-shot-down-iran-bravo-rescue/"],"source_refs":["AIBD-0328-source-0001","AIBD-0328-source-0002","AIBD-0328-source-0003","AIBD-0328-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"one rescued airman; the claimed AI contribution remains unverified","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"one rescued airman; the claimed AI contribution remains unverified","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0328-reported-measure","evaluation_data":null},{"case_id":"AIBD-0639","case_url":"records/AIBD-0639.html","measure":"Change in anxiety, depression, insomnia and tinnitus distress.","reported_result":"Woebot users lowered depression scores on the PHQ-9 from 10.03 to 8.06","baseline":"Waitlist PHQ-9 from 9.29 to 9.64 over eight weeks","result":"Woebot PHQ-9 from 10.03 to 8.06 over eight weeks","baseline_value":null,"result_value":null,"unit":null,"effect":"Woebot: PHQ-9 -1.97, GAD-7 -1.47; Tinnibot: PHQ-9 -3.09, GAD-7 -2.50, TFI -13.23; control no reliable change.","design":"randomized.","sample":"105 UK adults with self-reported persistent tinnitus, three arms of 36 with three withdrawals.","uncertainty":"Self-reported diagnosis and outcomes; trial not prospectively registered; small arms; post-intervention only.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC13097093"],"source_refs":["AIBD-0639-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Isolated against no treatment.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Waiting-list control.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Woebot: PHQ-9 -1.97, GAD-7 -1.47; Tinnibot: PHQ-9 -3.09, GAD-7 -2.50, TFI -13.23; control no reliable change.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Daily use for eight weeks.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"randomized.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Self-report GAD-7, PHQ-9, Insomnia Severity Index and Tinnitus Functional Index before and after eight weeks.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Change in anxiety, depression, insomnia and tinnitus distress.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Data available from the corresponding author on request.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"105 UK adults with self-reported persistent tinnitus, three arms of 36 with three withdrawals.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Self-reported diagnosis and outcomes; trial not prospectively registered; small arms; post-intervention only.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0639-reported-measure","evaluation_data":null},{"case_id":"AIBD-0235","case_url":"records/AIBD-0235.html","measure":null,"reported_result":null,"baseline":"officers may need to spend hours manually reviewing seized material to establish whether images or videos relate to known cases","result":"could significantly shorten the time between detection and intervention","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://news.met.police.uk/news/met-considers-ai-to-quickly-identify-child-sexual-abuse-victims-508131","https://oecd.ai/en/incidents/2026-04-13-83b8"],"source_refs":["AIBD-0235-source-0001","AIBD-0235-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: manual review of seized material by officers","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: officers may need to spend hours manually reviewing seized material to establish whether images or videos relate to known cases. Result: could significantly shorten the time between detection and intervention.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0235-reported-measure","evaluation_data":null},{"case_id":"AIBD-0374","case_url":"records/AIBD-0374.html","measure":"more than 22,000 missions on Ukraine's front line","reported_result":"more than 22,000 missions on Ukraine's front line","baseline":null,"result":"The position was taken and its defenders surrendered, with no Ukrainian losses, per the President","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.businessinsider.com/ukraine-russia-position-taken-without-using-troops-just-robots-drones-2026-4","https://www.politico.eu/article/volodymyr-zelenskyy-robotic-systems-russia-army-positions-ukraine/","https://censor.net/en/news/3610314/ukraine-s-defense-forces-seize-position-using-only-robots-and-drones","https://arstechnica.com/ai/2026/04/ukraines-military-robot-surge-aims-to-offset-drone-risks-to-humans/","https://www.newsweek.com/russia-ukraine-war-latest-putin-troops-surrender-military-robot-first-time-11825010","https://thedebrief.org/ukraine-says-it-captured-a-russian-position-using-only-unmanned-systems-a-glimpse-of-future-warfare/","https://theconversation.com/robots-just-captured-a-russian-position-in-ukraine-but-dont-worry-about-real-life-terminators-just-yet-280959","https://oecd.ai/en/incidents/2026-04-13-ec31"],"source_refs":["AIBD-0374-source-0001","AIBD-0374-source-0002","AIBD-0374-source-0003","AIBD-0374-source-0004","AIBD-0374-source-0005","AIBD-0374-source-0006","AIBD-0374-source-0007","AIBD-0374-source-0008"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"more than 22,000 missions on Ukraine's front line","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"more than 22,000 missions on Ukraine's front line","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0374-reported-measure","evaluation_data":null},{"case_id":"AIBD-0343","case_url":"records/AIBD-0343.html","measure":"released 175 tonnes of methane in a single week","reported_result":"released 175 tonnes of methane in a single week","baseline":"175 t of methane (4,900 tCO2e) in a single week in April","result":"Operators completed the repair work one week after the methane plume was first detected.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.thechemicalengineer.com/news/canada-plugs-major-source-of-methane-emissions-from-oil-and-gas-field-leak/","https://wedocs.unep.org/bitstreams/03e1cce3-cce5-4cbb-8574-c9a004b3df20/download","https://www.unep.org/topics/energy/methane/methane-alert-and-response-system-mars"],"source_refs":["AIBD-0343-source-0002","AIBD-0343-source-0003","AIBD-0343-source-0004","AIBD-0343-source-0005","AIBD-0343-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"the leak before repair","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 175 t of methane (4,900 tCO2e) in a single week in April. Result: Operators completed the repair work one week after the methane plume was first detected.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"released 175 t of methane (4,900 tCO2e) in a single week","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0343-reported-measure","evaluation_data":null},{"case_id":"AIBD-0262","case_url":"records/AIBD-0262.html","measure":"36.5% complete skin clearance with zasocitinib versus 13.9% with deucravacitinib at week 16","reported_result":"36.5% complete skin clearance with zasocitinib versus 13.9% with deucravacitinib at week 16","baseline":"13.9% of those receiving deucravacitinib (n = 303)","result":"36.5% of patients (n = 301) achieving PASI 100 at week 16","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://www.takeda.com/newsroom/newsreleases/2026/zasocitinib-superior-efficacy/","https://pmc.ncbi.nlm.nih.gov/articles/PMC11339701/","https://clinicaltrials.gov/study/NCT06973291","https://www.schrodinger.com/design-of-a-highly-selective-allosteric-picomolar-tyk2-inhibitor-using-novel-fep-strategies/","https://www.drugtargetreview.com/inside-zasocitinib-a-new-model-for-tyk2-inhibition-in-immune-mediated-diseases/677029.article","https://www.takeda.com/newsroom/newsreleases/2026/fda-priority-review-zasocitinib-psoriasis/","https://www.takeda.com/newsroom/newsreleases/2023/takeda-completes-acquisition-of-nimbus-therapeutics-tyk2-program-subsidiary/","https://pmc.ncbi.nlm.nih.gov/articles/PMC12327466/"],"source_refs":["AIBD-0262-source-0001","AIBD-0262-source-0002","AIBD-0262-source-0003","AIBD-0262-source-0004","AIBD-0262-source-0005","AIBD-0262-source-0006","AIBD-0262-source-0007"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"deucravacitinib (Sotyktu), the active comparator in the LATITUDE Atlas phase 3 trial","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: 13.9% of those receiving deucravacitinib (n = 303). Result: 36.5% of patients (n = 301) achieving PASI 100 at week 16.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"PASI 100 36.5% vs 13.9% for deucravacitinib at week 16","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0262-reported-measure","evaluation_data":null},{"case_id":"AIBD-0661","case_url":"records/AIBD-0661.html","measure":"Binding affinity and antagonist activity.","reported_result":"bound MRGPRX2 with affinities between about 20 and 200 nM","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":"Three binders at 20 to 200 nM that blocked degranulation.","design":"other.","sample":"10,000 in silico nanobodies; ten expressed and tested.","uncertainty":"One receptor; ten candidates; no direct structure of the complex; patent interest of some authors.","comparison_basis":null,"source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC13315262/"],"source_refs":["AIBD-0661-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not quantified against an experimental screen.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Negative control nanobody (Nb 60).","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Three binders at 20 to 200 nM that blocked degranulation.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Laboratory.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Flow cytometry binding, mast cell degranulation and G protein signaling assays; mutagenesis.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Binding affinity and antagonist activity.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Ranking code on GitHub; data on request.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"10,000 in silico nanobodies; ten expressed and tested.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"One receptor; ten candidates; no direct structure of the complex; patent interest of some authors.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0661-reported-measure","evaluation_data":null},{"case_id":"AIBD-0348","case_url":"records/AIBD-0348.html","measure":"496,000 hectares affected between 2018 and 2025","reported_result":"496,000 hectares affected between 2018 and 2025","baseline":null,"result":"specificity 0.976, sensitivity 0.991","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://github.com/earthrise-media/mining-detector","https://news.mongabay.com/2026/04/ai-tool-tracks-spread-of-illegal-gold-mining-in-amazon-protected-areas/","https://amazonminingwatch.org/en/about","https://www.maapprogram.org/lapampa_opermercury/","https://web.archive.org/web/2026/https://www.niemanlab.org/2026/04/geospatial-ai-is-reinventing-the-rainforest-beat/","https://latamjournalismreview.org/articles/journalists-in-africa-and-south-america-are-mapping-the-mines-devouring-the-worlds-rainforests/"],"source_refs":["AIBD-0348-source-0002","AIBD-0348-source-0003","AIBD-0348-source-0004","AIBD-0348-source-0001","AIBD-0348-source-0005","AIBD-0348-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"hand-labeled Sentinel-2 patches (Venezuela geographic holdout)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Result: specificity 0.976, sensitivity 0.991.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"496,000 hectares affected between 2018 and 2025","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0348-reported-measure","evaluation_data":null},{"case_id":"AIBD-0075","case_url":"records/AIBD-0075.html","measure":null,"reported_result":null,"baseline":"open (status last updated 31 August 2025)","result":"proved (Lean), updated 1 May 2026","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://raw.githubusercontent.com/teorth/erdosproblems/13b930b4be528278d778e6bb8c3fb590bd5ca05d/data/problems.yaml","https://www.erdosproblems.com/38","https://www.erdosproblems.com/forum/thread/38","https://github.com/teorth/erdosproblems/wiki/AI-contributions-to-Erd%C5%91s-problems","https://raw.githubusercontent.com/teorth/erdosproblems/b2de5c4b114ce03e01cdf43aa00f27ada4566757/data/problems.yaml"],"source_refs":["AIBD-0075-source-0001","AIBD-0075-source-0002","AIBD-0075-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0075-reported-measure","evaluation_data":null},{"case_id":"AIBD-0340","case_url":"records/AIBD-0340.html","measure":"over 800 leaks across the network, saving an estimated 8.7 million liters of water a day","reported_result":"over 800 leaks across the network, saving an estimated 8.7 million liters of water a day","baseline":"a person working unassisted on the ground","result":"at least five times quicker","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.bbc.co.uk/news/articles/cr5pzmyr5jeo","https://www.thameswater.co.uk/news/2026/apr/thames-water-signs-landmark-leak-detection-contract-with-origin-tech-to-save-over-100-million-litres-of-water-per-day","https://watermagazine.co.uk/2026/05/29/thames-water-signs-landmark-leak-detection-contract-with-origin-tech-to-save-over-100-million-litres-of-water-per-day/","https://smartwatermagazine.com/news/thames-water/thames-water-signs-ai-satellite-deal-plug-100-million-litres-daily-leaks","https://www.origintech.co.uk/origin-news/origin-tech-in-partnership-with-thames-water","https://www.origintech.co.uk/origin-news/origin-tech-featured-in-bbc-news-and-bbc-wiltshire-radio","https://www.thameswater.co.uk/news/2026/aug/thames-water-uses-satellites-ai-and-thousands-of-acoustic-sensors-to-hunt-down-leaks-as-south-east-drought-pressure-intensifies","https://www.origintech.co.uk/origin-news/origin-technologies-make-the-times"],"source_refs":["AIBD-0340-source-0002","AIBD-0340-source-0003","AIBD-0340-source-0004","AIBD-0340-source-0005","AIBD-0340-source-0006","AIBD-0340-source-0007","AIBD-0340-source-0008","AIBD-0340-source-0001","AIBD-0340-source-0009"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: a person detecting leaks unassisted on the ground (the vendor co-founder's claim)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: a person working unassisted on the ground. Result: at least five times quicker.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"over 800 leaks across the network, saving an estimated 8.7 million liters of water a day","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0340-reported-measure","evaluation_data":null},{"case_id":"AIBD-0226","case_url":"records/AIBD-0226.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://fox49.tv/news/local/florida-state-uses-zeroeyes-tech-to-detect-guns-on-campus-police-say","https://www.wctv.tv/2026/05/04/court-documents-19-year-old-arrested-after-brandishing-weapon-fsu-parking-garage-early-saturday-morning/","https://www.tallahassee.com/story/news/local/2026/05/05/ai-tool-identifying-firearm-leads-to-arrest-of-fsu-student/89946929007/","https://undark.org/2026/02/13/as-more-schools-turn-to-ai-weapons-detection-questions-persist/"],"source_refs":["AIBD-0226-source-0001","AIBD-0226-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0226-reported-measure","evaluation_data":null},{"case_id":"AIBD-0257","case_url":"records/AIBD-0257.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.skai.gr/news/politics/kyranakis-ston-skai-systima-ai-tha-elegxei-tis-synomilies-sta-trena"],"source_refs":["AIBD-0257-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0257-reported-measure","evaluation_data":null},{"case_id":"AIBD-0252","case_url":"records/AIBD-0252.html","measure":"detections had passed 16,900 by early July","reported_result":"detections had passed 16,900 by early July","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://news.ucsb.edu/2026/022586/new-ai-system-uses-cameras-and-thermal-sensors-steer-ships-clear-gray-whales-san","https://spectrum.ieee.org/whales-ai-thermal-camera-tracking","https://baynature.org/2026/07/02/science-nature/urban-nature/saving-the-whales-just-got-a-little-easier/","https://web.archive.org/web/20260727150147/https://insideclimatenews.org/news/26072026/san-francisco-ai-whale-technology/","https://www.cbsnews.com/sanfrancisco/news/san-francisco-bay-whale-stranding-13-pier-80-suspected-vessel-strike/"],"source_refs":["AIBD-0252-source-0001","AIBD-0252-source-0002","AIBD-0252-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0252-reported-measure","evaluation_data":null},{"case_id":"AIBD-0300","case_url":"records/AIBD-0300.html","measure":null,"reported_result":null,"baseline":"AF paroxysms were not documented by ambulatory monitoring or standard ECG","result":"AF paroxysms were detected repeatedly in the morning after notification of nocturnal “elevated” breathing disturbance, but not in the morning after nights when breathing disturbance was “not elevated”","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC13301138/","https://www.accessdata.fda.gov/cdrh_docs/pdf24/K240929.pdf"],"source_refs":["AIBD-0300-source-0001","AIBD-0300-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"ambulatory monitoring and standard 12-lead ECG, which had documented no paroxysms since his retirement","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Reference: AF paroxysms were not documented by ambulatory monitoring or standard ECG. Result: AF paroxysms were detected repeatedly in the morning after notification of nocturnal “elevated” breathing disturbance, but not in the morning after nights when breathing disturbance was “not elevated”.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0300-reported-measure","evaluation_data":null},{"case_id":"AIBD-0260","case_url":"records/AIBD-0260.html","measure":"eight years of probation, two years of GPS monitoring","reported_result":"eight years of probation, two years of GPS monitoring","baseline":"no report to law enforcement; the threats came from burner numbers","result":"detectives matched the anonymous texts to Zhou through the ChatGPT records and arrested him six days after the FBI notice","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.sun-sentinel.com/2026/08/14/financial-analyst-sentenced-to-probation-after-telling-chatgpt-about-plans-to-rape-and-kill-his-ex/","https://bocanewsnow.com/2026/08/27/man-flagged-by-chatgpt-pleads-guilty-in-stalking-case/","https://www3.pbso.org/blotter/","https://openai.com/index/helping-people-when-they-need-it-most/"],"source_refs":["AIBD-0260-source-0001","AIBD-0260-source-0002","AIBD-0260-source-0003","AIBD-0260-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"arrest on May 29, 2026; guilty plea on August 13","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"arrest on May 29, 2026; guilty plea on August 13","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0260-reported-measure","evaluation_data":null},{"case_id":"AIBD-0234","case_url":"records/AIBD-0234.html","measure":"two people at apparent risk of suicide","reported_result":"two people at apparent risk of suicide","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.straitstimes.com/asia/east-asia/japan-stations-facilities-use-ai-system-to-prevent-suicide-by-jumping","https://jp.asilla.com/post/version-2-4-20240126"],"source_refs":["AIBD-0234-source-0001","AIBD-0234-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"two reported interventions after alerts; dates and facilities undisclosed","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"two reported interventions after alerts; dates and facilities undisclosed","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0234-reported-measure","evaluation_data":null},{"case_id":"AIBD-0330","case_url":"records/AIBD-0330.html","measure":null,"reported_result":null,"baseline":"a fault in the prototype after a communication glitch, in Qualcomm's account","result":"a designed safe collapse to the knees, in Qualcomm's account","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.creativebloq.com/design/product-design/this-video-of-an-ai-robot-collapsing-mid-presentation-gets-funnier-every-time-i-watch-it","https://oecd.ai/en/incidents/2026-07-28-5a37","https://www.ndtv.com/offbeat/watch-ai-humanoid-robot-collapses-in-taiwan-during-live-demo-video-viral-11845190","https://www.ndtvprofit.com/technology/humanoid-robot-running-on-qualcomm-ai-chip-dies-on-stage-at-computex-2026-11844305","https://www.techspot.com/news/113313-qualcomm-powered-humanoid-robot-collapses-stage-like-shot.html","https://www.deccanchronicle.com/technology/humanoid-robot-powered-by-qualcomm-collapses-on-stage-at-computex-2026-1975284","https://newsable.asianetnews.com/world/taiwan-ai-humanoid-robot-crashes-onstage-during-live-demo-at-computex-2026-articleshow-bcr0em1","https://www.gadgetreview.com/a-humanoid-robot-dies-on-stage-at-computex-2026-gets-carried-off","https://events.computextaipei.com.tw/en/","https://neura-robotics.com/products/4ne1"],"source_refs":["AIBD-0330-source-0001","AIBD-0330-source-0002","AIBD-0330-source-0003","AIBD-0330-source-0004","AIBD-0330-source-0005","AIBD-0330-source-0006","AIBD-0330-source-0007","AIBD-0330-source-0008","AIBD-0330-source-0009","AIBD-0330-source-0010"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0330-reported-measure","evaluation_data":null},{"case_id":"AIBD-0208","case_url":"records/AIBD-0208.html","measure":"she could identify E optotypes (acuity 0.03, peak 0.1)","reported_result":"she could identify E optotypes (acuity 0.03, peak 0.1)","baseline":"60 channels (international products of the same kind)","result":"256-channel flexible electrode array, with channel count and visual resolution reported as significantly superior","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://news.china.com/socialgd/10000169/20260606/49533586.html","https://news.sciencenet.cn/htmlnews/2026/6/566108.shtm","https://hunan.voc.com.cn/news/202606/32846000.html","https://www.xkb.com.cn/articleDetail/499962","https://www.21jingji.com/article/20260608/herald/908d9a3949ada35c98e6b666d4b41bc0.html","https://www.3dmgame.com/news/202606/3945762.html","https://finance.ifeng.com/c/8tmrCoDSlTF","https://doi.org/10.1167/tvst.10.10.14","https://pmc.ncbi.nlm.nih.gov/articles/PMC13615219/"],"source_refs":["AIBD-0208-source-0001","AIBD-0208-source-0002","AIBD-0208-source-0003","AIBD-0208-source-0004","AIBD-0208-source-0005","AIBD-0208-source-0006","AIBD-0208-source-0007","AIBD-0208-source-0008"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: international retinal prostheses of the same type with 60 channels","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: 60 channels (international products of the same kind). Result: 256-channel flexible electrode array, with channel count and visual resolution reported as significantly superior.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"she could identify E optotypes (acuity 0.03, peak 0.1)","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0208-reported-measure","evaluation_data":null},{"case_id":"AIBD-0010","case_url":"records/AIBD-0010.html","measure":"a Sardhana team reached his home within eight minutes","reported_result":"a Sardhana team reached his home within eight minutes","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.deccanherald.com/india/uttar-pradesh/when-every-minute-counts-uttar-pradesh-police-save-youth-attempting-suicide-in-8-minutes-after-meta-alert-over-instagram-post-4032393","https://www.thehansindia.com/news/national/cops-save-man-in-8-mins-after-suicide-alert-on-insta-1084758","https://www.northcountrypublicradio.org/news/npr/668408122/facebook-increasingly-reliant-on-a-i-to-predict-suicide-risk","https://about.fb.com/news/2018/09/inside-feed-suicide-prevention-and-ai/","https://www.tribuneindia.com/news/nation/facebook-helps-up-police-avert-two-suicides-in-a-week-in-state-476915/amp","https://www.tribuneindia.com/news/delhi/man-saved-by-delhi-police-after-facebook-alerts-on-self-harm-video-263850/amp","https://www.decodeinternet.in/decode/meta-sent-suicide-alert-uttar-pradesh-police-had-minutes-to-save-30394"],"source_refs":["AIBD-0010-source-0001","AIBD-0010-source-0002","AIBD-0010-source-0003","AIBD-0010-source-0004","AIBD-0010-source-0005","AIBD-0010-source-0006"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"3,011 lives saved statewide from 1 January 2023 to 31 May 2026","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"3,011 lives saved statewide from 1 January 2023 to 31 May 2026","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0010-reported-measure","evaluation_data":null},{"case_id":"AIBD-0209","case_url":"records/AIBD-0209.html","measure":"both were rescued within approximately two hours","reported_result":"both were rescued within approximately two hours","baseline":"crewed rescue craft put more personnel at risk; during the April rescue an A-10 was hit by anti-aircraft fire","result":"an unmanned vessel reached and carried the two soldiers, and a helicopter completed the recovery in safer waters","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.stripes.com/branches/navy/2026-06-10/sea-drones-us-navy-deployments-21929112.html","https://www.dvidshub.net/news/567214/us-army-crew-safely-rescued-after-helicopter-lost-sea","https://defensescoop.com/2026/06/09/autonomous-maritime-drone-rescues-apache-pilots-after-crash/","https://www.al-monitor.com/originals/2026/06/drone-rescue-highlights-us-navys-autonomous-push","https://www.stripes.com/theaters/us/2026-09-18/hegseth-austin-arsenal-freedom-tech-22892957.html"],"source_refs":["AIBD-0209-source-0001","AIBD-0209-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"both were rescued within approximately two hours","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"both were rescued within approximately two hours","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0209-reported-measure","evaluation_data":null},{"case_id":"AIBD-0254","case_url":"records/AIBD-0254.html","measure":null,"reported_result":null,"baseline":"an average of five trespass incidents a day","result":"incidents have reduced (no figure given)","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.networkrailmediacentre.co.uk/news/ai-cameras-drive-down-dingwall-trespass-but-safety-message-remains"],"source_refs":["AIBD-0254-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: trespass recorded at the same crossings before the cameras (June 2025 monitoring)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: an average of five trespass incidents a day. Result: incidents have reduced (no figure given).","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0254-reported-measure","evaluation_data":null},{"case_id":"AIBD-0602","case_url":"records/AIBD-0602.html","measure":"Rate of collisions with pedestrians or cyclists, and with motorcycles.","reported_result":"a 17% lower rate of pedestrian or cyclist collisions (95% CI 5% to 27%)","baseline":"Collision rate with pedestrians or cyclists of vehicles without AEB","result":"17% lower rate (95% CI 5% to 27%)","baseline_value":null,"result_value":null,"unit":null,"effect":"17% lower pedestrian or cyclist collision rate (95% CI 5% to 27%); 11% lower motorcycle collision rate (95% CI 0% to 21%).","design":"controlled observational.","sample":"Australian and New Zealand injury crashes, 2016 to 2023.","uncertainty":"Observational; fitment inferred from Redbook specifications; too few cyclists for a separate estimate.","comparison_basis":"measured","source_urls":["https://pubmed.ncbi.nlm.nih.gov/42275024/"],"source_refs":["AIBD-0602-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"No reduction for AEB without pedestrian detection, so the gain is tied to the detection.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Vehicles without AEB; vehicles with AEB lacking pedestrian detection.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"17% lower pedestrian or cyclist collision rate (95% CI 5% to 27%); 11% lower motorcycle collision rate (95% CI 0% to 21%).","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"Vehicles listed with pedestrian avoidance with braking.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Induced exposure analysis of injury crash rates, controlling for jurisdiction, speed limit, market group, driver age and sex, weather, light and year.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Rate of collisions with pedestrians or cyclists, and with motorcycles.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Jurisdictional crash data and Redbook specifications; code not public.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Australian and New Zealand injury crashes, 2016 to 2023.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Observational; fitment inferred from Redbook specifications; too few cyclists for a separate estimate.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0602-reported-measure","evaluation_data":null},{"case_id":"AIBD-0375","case_url":"records/AIBD-0375.html","measure":"identified all 18 faces in about 30 minutes","reported_result":"identified all 18 faces in about 30 minutes","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://tuoitre.vn/tu-camera-ai-cong-an-chi-mat-30-phut-nhan-dien-18-thanh-thieu-nien-tu-tap-danh-nhau-tren-pho-20260614143550899.htm","https://vietnamnet.vn/18-thanh-thieu-nien-gay-roi-bi-camera-ai-o-phu-tho-nhan-dien-sau-30-phut-2525799.html","https://tienphong.vn/camera-ai-tom-gon-nhom-hoc-sinh-tu-tap-mang-theo-hung-khi-post1851303.tpo","https://oecd.ai/en/incidents/2026-06-14-ba55","https://thanhnien.vn/khoi-to-10-thanh-thieu-nien-hon-chien-cong-khai-danh-tinh-cha-me-tung-nguoi-185260619214040298.htm"],"source_refs":["AIBD-0375-source-0001","AIBD-0375-source-0002","AIBD-0375-source-0003","AIBD-0375-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"identified all 18 faces in about 30 minutes","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"identified all 18 faces in about 30 minutes","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0375-reported-measure","evaluation_data":null},{"case_id":"AIBD-0210","case_url":"records/AIBD-0210.html","measure":"retirement savings of more than $3 million","reported_result":"retirement savings of more than $3 million","baseline":"a request to withdraw the whole retirement portfolio of about $3 million","result":"the money did not move","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://fortune.com/2026/06/17/tiaa-saved-retiree-from-scam-using-artifical-intelligence-human-centered-work-ceo-thasunda-brown-duckett/","https://fortune.com/2026/06/22/tiaa-ceo-thasunda-brown-duckett-retirement-crisis-annuity-income-america/","https://web.archive.org/web/2026/https://www.forbes.com/sites/brucelee/2026/03/08/how-tiaa-is-using-ai-to-help-protect-older-adults/"],"source_refs":["AIBD-0210-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"retirement savings above $3 million retained, according to TIAA’s CEO","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"retirement savings above $3 million retained, according to TIAA’s CEO","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0210-reported-measure","evaluation_data":null},{"case_id":"AIBD-0239","case_url":"records/AIBD-0239.html","measure":"held him a day before his stated date","reported_result":"held him a day before his stated date","baseline":"Police received the tip on 16 June 2026.","result":"Police arrested him on 19 June 2026, a day before that date.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://pc.es.gov.br/Not%C3%ADcia/alerta-do-fbi-e-do-ministerio-da-justica-auxiliam-pces-na-prisao-de-investigado-por-ameacas-contra-o-proprio-filho-em-inteligencia-artificial","https://g1.globo.com/es/espirito-santo/norte-noroeste-es/noticia/2026/06/25/pai-e-preso-por-planejar-a-morte-do-filho-para-nao-pagar-pensao-e-confessar-crime-ao-chatgpt.ghtml","https://www.agazeta.com.br/es/policia/conversa-com-ia-ajuda-policia-a-prender-pai-por-planejar-morte-de-filho-no-es-0626","https://www1.folha.uol.com.br/cotidiano/2026/06/homem-e-preso-sob-suspeita-de-usar-ia-para-planejar-morte-de-filho-de-8-anos-plataforma-emitiu-alerta.shtml","https://g1.globo.com/tecnologia/noticia/2026/06/27/como-chatgpt-identificou-plano-de-pai-para-matar-o-filho-e-levou-alerta-para-autoridades.ghtml","https://www.bbc.com/portuguese/articles/c8dn1l7vr6ro","https://openai.com/index/helping-people-when-they-need-it-most/"],"source_refs":["AIBD-0239-source-0001","AIBD-0239-source-0002","AIBD-0239-source-0003","AIBD-0239-source-0004","AIBD-0239-source-0005","AIBD-0239-source-0006","AIBD-0239-source-0007"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"held him a day before his stated date","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"held him a day before his stated date","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0239-reported-measure","evaluation_data":null},{"case_id":"AIBD-0012","case_url":"records/AIBD-0012.html","measure":"located less than five hours after the report","reported_result":"located less than five hours after the report","baseline":"several days (the inspector's estimate of the search time avoided)","result":"less than five hours after they were reported missing","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.abc.net.au/news/2026-06-27/lost-hikers-found-by-ai-drone-in-kosciuszko-national-park/106846712","https://www.fire.nsw.gov.au/media/news/2026/20260625-fire-and-rescue-nsw-drone-team-locates-missing-hikers-using-ai-technology","https://aboutregional.com.au/how-ai-saved-two-hikers-lost-in-the-unforgiving-alpine-wilderness/509715/","https://itwire.com/business-it-news/business-technology/ai-assisted-drone-helps-rescue-lost-kosciuszko-hikers-in-nsw-first"],"source_refs":["AIBD-0012-source-0002","AIBD-0012-source-0003","AIBD-0012-source-0004"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: a conventional search without the AI drone","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: several days (the inspector's estimate of the search time avoided). Result: less than five hours after they were reported missing.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"located less than five hours after the report","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0012-reported-measure","evaluation_data":null},{"case_id":"AIBD-0246","case_url":"records/AIBD-0246.html","measure":"trains made 17 emergency stops on detected anomalies in the system's first two months","reported_result":"trains made 17 emergency stops on detected anomalies in the system's first two months","baseline":"People left inside a crossing are noticed by the driver's sight or the emergency button","result":"The AI sends a stop signal to approaching trains; 17 emergency stops in the first two months","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.yomiuri.co.jp/national/20260919-GYT1T00229/2/","https://www.itmedia.co.jp/news/articles/2606/22/news125.html","https://railf.jp/news/2026/06/25/041500.html","https://oecd.ai/en/incidents/2026-06-22-b4f8","https://www.toyotasystems.com/news/upload/2026/06/9230e25b62c7992c0aa938759e7a7fd9589f2be2.pdf","https://www.yomiuri.co.jp/national/20260919-GYT1T00229/3/"],"source_refs":["AIBD-0246-source-0001","AIBD-0246-source-0002","AIBD-0246-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0246-reported-measure","evaluation_data":null},{"case_id":"AIBD-0360","case_url":"records/AIBD-0360.html","measure":null,"reported_result":null,"baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://aboutregional.com.au/ai-cameras-to-be-extra-set-of-eyes-on-roadside-wildlife-in-monaro-highway-trial/510964/","https://www.nsw.gov.au/ministerial-releases/new-ai-animal-detection-and-smart-signage-trial-working-to-limit-wildlife-vehicle-strikes"],"source_refs":["AIBD-0360-source-0001","AIBD-0360-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0360-reported-measure","evaluation_data":null},{"case_id":"AIBD-0292","case_url":"records/AIBD-0292.html","measure":null,"reported_result":null,"baseline":"it was recommended that the patient undergo continued cardiac monitoring for 30 days to attempt to document the arrhythmia","result":"his care was able to be expedited straight to discussing treatment options","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://www.acofp.org/docs/default-source/ofp/vol-18-no-3-summer-2026/11_consumer-wearable-device-leads-to-expedited-diagnosis-and-treatment-of-paroxysmal-atrial-fibrillation.pdf"],"source_refs":["AIBD-0292-source-0001"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: repeat 30-day ambulatory cardiac monitoring","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: it was recommended that the patient undergo continued cardiac monitoring for 30 days to attempt to document the arrhythmia. Result: his care was able to be expedited straight to discussing treatment options.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0292-reported-measure","evaluation_data":null},{"case_id":"AIBD-0376","case_url":"records/AIBD-0376.html","measure":"three arrests and six Sexual Harm Prevention Order checks","reported_result":"three arrests and six Sexual Harm Prevention Order checks","baseline":null,"result":null,"baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":null,"source_urls":["https://www.westyorkshire.police.uk/news-appeals/cutting-edge-facial-recognition-cameras-used-first-time-kirklees","https://www.westyorkshire.police.uk/sites/default/files/2026-10/live-facial-recognition-deployment-register-2026-19102026.docx","https://www.westyorkshire.police.uk/about-us/how-we-work/facial-recognition/live-facial-recognition","https://science.police.uk/site/assets/files/3396/frt-equitability-study_mar2023.pdf"],"source_refs":["AIBD-0376-source-0001","AIBD-0376-source-0002","AIBD-0376-source-0003"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"three arrests and six Sexual Harm Prevention Order stops","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"three arrests and six Sexual Harm Prevention Order stops","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0376-reported-measure","evaluation_data":null},{"case_id":"AIBD-0598","case_url":"records/AIBD-0598.html","measure":null,"reported_result":null,"baseline":"Staff reviewed cases one by one to find patterns and asked the existing vendor for custom reports","result":"The trends and mapping pages showed request locations and common case types automatically","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://get-civic.com/articles/what-team-moulton-taught-us-about-building-govtech-right","https://get-civic.com/","https://www.house.gov/sites/default/files/2026-08/grids/APRIL-JUNE%202026%20SOD%20DETAIL%20GRID-FINAL.csv"],"source_refs":["AIBD-0598-source-0001","AIBD-0598-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The comparison is not established.","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The size of the effect is not established.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The outcome measure is not established.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0598-reported-measure","evaluation_data":null},{"case_id":"AIBD-0283","case_url":"records/AIBD-0283.html","measure":"Number and fraction of confident catalog events recovered, by mass bin.","reported_result":"47 detections with a false alarm rate below one per year and p_astro above 0.9","baseline":null,"result":"47 detections, 17% of the catalog population and 39% of the cWB count","baseline_value":null,"result_value":null,"unit":null,"effect":"47 detections, 17% of the catalog population and 39% of the coherent WaveBurst count; 44 shared with coherent WaveBurst; no candidate without a catalog counterpart.","design":"other.","sample":"LIGO Hanford and Livingston data from O3a, O3b, O4a and O4b, 442.9 days of coincident livetime.","uncertainty":"Written and run by the pipeline's developers; not peer reviewed; recovery falls to 8.1% of confident events below 30 solar masses.","comparison_basis":"measured","source_urls":["https://arxiv.org/html/2609.39583v2","https://arxiv.org/abs/2609.39583v1","https://github.com/ginguglia/MADGRAV/tree/acb66841f9408939c1ca17bff90360214c4ea2c9"],"source_refs":["https://arxiv.org/abs/2609.39583v2"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"None over coherent WaveBurst: the detections are nearly a subset of its events.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Published gravitational-wave catalogs and the coherent WaveBurst search on the same data.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"47 detections, 17% of the catalog population and 39% of the coherent WaveBurst count; 44 shared with coherent WaveBurst; no candidate without a catalog counterpart.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Retrospective search of archival data; detections are counted at a calibrated false alarm rate below one per year with p_astro above 0.9 and matched to published catalog events and to coherent WaveBurst results.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Number and fraction of confident catalog events recovered, by mass bin.","inferred":false}},"evaluations[].reproduction_materials":{"state":"value_present","review_state":"reviewed","display":{"answer":"Code, trained weights and frozen calibration in the public GitHub repository.","inferred":false}},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"LIGO Hanford and Livingston data from O3a, O3b, O4a and O4b, 442.9 days of coincident livetime.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Written and run by the pipeline's developers; not peer reviewed; recovery falls to 8.1% of confident events below 30 solar masses.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0283-reported-measure","evaluation_data":null},{"case_id":"AIBD-0359","case_url":"records/AIBD-0359.html","measure":"Average change in speed as vehicles approached activated signs; overall detection accuracy for medium and large wildlife.","reported_result":"an average speed reduction of 3 mph as vehicles approached activated signs","baseline":null,"result":"an average speed reduction of 3 mph as vehicles approached signs that had been activated by wildlife detections","baseline_value":null,"result_value":null,"unit":null,"effect":"An average speed reduction of 3 mph at activated signs and 97 percent detection accuracy.","design":"other.","sample":"More than 10,000 passing vehicles on Route 460 in Southwest Virginia, where eight trailers covered a 750-meter section, during a one-year pilot that ended in September 2026.","uncertainty":"Initial results from the evaluating institute. The report gives no baseline, no confidence interval and no crash outcome, and it states that data analysis is continuing.","comparison_basis":"measured","source_urls":["https://news.vt.edu/articles/2026/10/vtti-elk-live-driver-warning-detection-virginia.html","https://wcyb.com/news/local/vdot-tests-elk-detection-system-to-prevent-collisions-in-southwest-virginia","https://www.vdot.virginia.gov/media/vdotvirginiagov/projects/bristol/corridor-q-driver-safety-and-maintaining-habitat/VTRC-Evaluation-of-Wildlife-Crossing-Design-for-Corridor-Q-Final-Report_acc09032025_PM.pdf","https://www.vdot.virginia.gov/projects/bristol-district/corridor-q-driver-safety-and-maintaining-habitat-connectivity/"],"source_refs":["AIBD-0359-source-0001"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"reviewed"},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not stated in the report.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"An average speed reduction of 3 mph at activated signs and 97 percent detection accuracy.","inferred":false}},"evaluations[].exposure":{"state":"value_present","review_state":"reviewed","display":{"answer":"One year of operation.","inferred":false}},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"other.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Roadside radar units measured vehicle speeds along the corridor. The report gives an overall detection accuracy for medium and large wildlife and does not say how it was scored.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Average change in speed as vehicles approached activated signs; overall detection accuracy for medium and large wildlife.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"More than 10,000 passing vehicles on Route 460 in Southwest Virginia, where eight trailers covered a 750-meter section, during a one-year pilot that ended in September 2026.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Initial results from the evaluating institute. The report gives no baseline, no confidence interval and no crash outcome, and it states that data analysis is continuing.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0359-reported-measure","evaluation_data":null},{"case_id":"AIBD-0303","case_url":"records/AIBD-0303.html","measure":"About ten days later he returned for open-heart surgery","reported_result":"About ten days later he returned for open-heart surgery","baseline":"He probably would have had a stroke, heart attack, or both","result":"He says an ER doctor reviewed the tracing and confirmed that he was in AFib.","baseline_value":null,"result_value":null,"unit":null,"effect":null,"design":null,"sample":null,"uncertainty":null,"comparison_basis":"assumed","source_urls":["https://cyberguy.com/future-tech/galaxy-watch-afib-alert-dad-heart-surgery/","https://image-us.samsung.com/SamsungUS/home/apps/samsung-health-monitor/IFU_IHRN_US_202306.pdf","https://pmc.ncbi.nlm.nih.gov/articles/PMC8854369/"],"source_refs":["AIBD-0303-source-0001","AIBD-0303-source-0002"],"review_state":"not_assessed","field_qualifications":{"evaluations[].ai_uplift":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The improvement attributable to AI is not established.","inferred":false}},"evaluations[].comparator":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed alternative: no watch ECG reading (Joseph's and his doctors' counterfactual)","inferred":false}},"evaluations[].design":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The study design is unclear.","inferred":false}},"evaluations[].effect":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"Assumed comparison. Reference: He probably would have had a stroke, heart attack, or both. Result: He says an ER doctor reviewed the tracing and confirmed that he was in AFib.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The number of people or uses exposed to the system is unclear.","inferred":false}},"evaluations[].id":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"A specific evaluation has not been identified.","inferred":false}},"evaluations[].measurement_method":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"How the result was measured is unclear.","inferred":false}},"evaluations[].metric":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"open-heart surgery followed ten days later","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The availability of materials for replication is unclear.","inferred":false}},"evaluations[].setting_sample":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The evaluation setting and sample are unclear.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"unresolved","review_state":"not_assessed","display":{"answer":"The uncertainty around the result is not established.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0303-reported-measure","evaluation_data":null},{"case_id":"AIBD-0361","case_url":"records/AIBD-0361.html","measure":"Estimated cover of six benthic groups, including total hard coral cover.","reported_result":"Estimates of non-Acropora hard corals and encrusting soft corals were comparable across methods; ReefScan underestimated total hard coral cover relative to diver photo-transects. Traditional Owners chose ReefScan for regional assessments.","baseline":"30 points on each photoquadrat annotated by a single researcher","result":"20 points on each photoquadrat annotated by ReefCloud's model, with low consistency against the human annotations","baseline_value":null,"result_value":null,"unit":null,"effect":"Estimates of non-Acropora hard corals and encrusting soft corals were comparable across methods; ReefScan underestimated total hard coral cover relative to diver photo-transects. Traditional Owners chose ReefScan for regional assessments.","design":"controlled observational.","sample":"Reefs around Erub, Ugar and Mer islands, Kemer Kemer Meriam nation, eastern Torres Strait, Australia.","uncertainty":"Preprint, not peer reviewed; several authors are AIMS staff; one region.","comparison_basis":"measured","source_urls":["https://pmc.ncbi.nlm.nih.gov/articles/PMC12915753/","https://www.aims.gov.au/research/technology/reefscan","https://www.aims.gov.au/research/technology/reefcloud","https://reefcloud.ai/","https://doi.org/10.22541/authorea.15007672/v1","https://meetingorganizer.copernicus.org/OOS2025/OOS2025-941.html","https://pmc.ncbi.nlm.nih.gov/articles/PMC12327664/"],"source_refs":["AIBD-0361-source-0005"],"review_state":"reviewed","field_qualifications":{"evaluations[].ai_uplift":{"state":"value_present","review_state":"reviewed","display":{"answer":"Not isolated: all three methods used machine-learning image analysis, so the comparison concerns the collection platform.","inferred":false}},"evaluations[].comparator":{"state":"value_present","review_state":"reviewed","display":{"answer":"Diver photo-transects.","inferred":false}},"evaluations[].design":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].effect":{"state":"value_present","review_state":"reviewed","display":{"answer":"Estimates of non-Acropora hard corals and encrusting soft corals were comparable across methods; ReefScan underestimated total hard coral cover relative to diver photo-transects. Traditional Owners chose ReefScan for regional assessments.","inferred":false}},"evaluations[].exposure":{"state":"unresolved","review_state":"reviewed"},"evaluations[].id":{"state":"value_present","review_state":"reviewed","display":{"answer":"controlled observational.","inferred":false}},"evaluations[].measurement_method":{"state":"value_present","review_state":"reviewed","display":{"answer":"Images from SCUBA photo-transects, vessel-deployed drop cameras and transom-mounted ReefScan units were analyzed with machine-learning models, and the methods were compared statistically on the cover of six benthic groups. Traditional Owners gave structured feedback on each method.","inferred":false}},"evaluations[].metric":{"state":"value_present","review_state":"reviewed","display":{"answer":"Estimated cover of six benthic groups, including total hard coral cover.","inferred":false}},"evaluations[].reproduction_materials":{"state":"unresolved","review_state":"reviewed"},"evaluations[].setting_sample":{"state":"value_present","review_state":"reviewed","display":{"answer":"Reefs around Erub, Ugar and Mer islands, Kemer Kemer Meriam nation, eastern Torres Strait, Australia.","inferred":false}},"evaluations[].uncertainty_limitations":{"state":"value_present","review_state":"reviewed","display":{"answer":"Preprint, not peer reviewed; several authors are AIMS staff; one region.","inferred":false}}},"extraction_basis":"Existing case information; no new source review implied.","observation_id":"AIBD-0361-reported-measure","evaluation_data":null}]