RapiD AI-assisted mapping up to 12.4% slower, then disabled

AI system

System: RapiD editor (AI-assisted mapping) used on HOT Tasking Manager projects; the sources do not name its developer
Model: Not assessed
Release: Not assessed
Provider: Not assessed
Role: Not assessed
Configuration: Not assessed
Inputs and outputs: Not assessed
Autonomy: Not assessed
Workflow: A HOT activation with large Tasking Manager project areas and Maxar imagery. RapiD AI-assisted editing offered to volunteer contributors. Validators reviewing contributions and feeding back on quality. Independent researchers comparing 1,650 contributors before and during the response.
Human oversight: Not assessed
Relation to AI: The AI component (RapiD AI-assisted mapping) is documented by the community-authored OSM wiki as counter-productive. It was up to 12.4% slower at task mapping, with validator-reported quality problems, and was then disabled on all activation projects. Use of OSM data by AKUT, AFAD, a tent-city health team and the World Bank is documented on the same page. It is not tied to AI-derived data. HOT's activation and size-up pages, from the Internet Archive, contain no AI mention. No cited source documents Microsoft building footprints, fAIr or Meta maps in the 2023 Türkiye, 2022 Pakistan or 2019 Idai responses.
AI contribution: The AI component (RapiD AI-assisted mapping) is documented by the community-authored OSM wiki as counter-productive. It was up to 12.4% slower at task mapping, with validator-reported quality problems, and was then disabled on all activation projects. Use of OSM data by AKUT, AFAD, a tent-city health team and the World Bank is documented on the same page. It is not tied to AI-derived data. HOT's activation and size-up pages, from the Internet Archive, contain no AI mention. No cited source documents Microsoft building footprints, fAIr or Meta maps in the 2023 Türkiye, 2022 Pakistan or 2019 Idai responses.

Benefits

Outcome identifiers: Not assessed
Benefit: contributors using RapiD mapped tasks up to 12.4% slower
Category: Reduced Task Time
Status: Not assessed
Beneficiary groups: People. Organizations. Society.
Beneficiaries: Not assessed
Reach: Not assessed
Distribution and access: Not assessed
Duration: Not assessed
Expected?: Yes
Costs and harms: Not assessed

Measurement

Studies: Not assessed
Study design: Not assessed
Measurement: Not assessed
Setting and sample: Not assessed
Measures: contributors using RapiD mapped tasks up to 12.4% slower
Comparison: Not assessed
Results: contributors using RapiD mapped tasks up to 12.4% slower
Additional value from AI: Not assessed
Limitations: Not assessed
Denominators: Not assessed
Materials for replication: Not assessed

Money

Kind of value: Not assessed
Amount: Not assessed
Currency and prices: Not assessed
Period and unit: Not assessed
Realization: Not assessed
Recipients and payers: Not assessed
Gross or net: Not assessed
Included costs: Not assessed
Comparison: Not assessed
Method: Not assessed
Return and payback: Not assessed

People and places

Participants: Operator: Humanitarian OpenStreetMap Team with Yer Çizenler (activation lead).
Jurisdictions: Benefit: Türkiye (Kahramanmaraş earthquakes) and northern Syria. Operator: United States.
Ownership and control: Not assessed
Dependencies: Not assessed
Cross-border connections: Not assessed
Geopolitical implications: Not assessed

Follow-up

Conditions for replication: Volunteer contributors of mixed experience using AI suggestions without extra validation. AI-suggested features (for example roads defaulting to “residential”) not corrected before upload. Validator capacity to detect and report quality problems.
Replication attempts: 1 follow-up entries. Outcome unresolved: 1.
Later findings: No later operational use or outcome is established.
Availability: Not assessed
Institutional decision: Not assessed
Grounds: Not assessed
Action and result: Not assessed
Success criterion: Not assessed

Evidence

Source numbers: AIBD-0353-source-0001–0005
Citations: Numbered references below.
Language: Not assessed
Source relationships: Operator
Original evidence: Not assessed
Access and versions: Not assessed
Supported findings: Not assessed
Evidence links: Not assessed
Evidence strength: Not assessed
Missing evidence: Not assessed

Grades

Assessed benefits: People. Organizations. Society.
Depth: Not assessed
Reasons: People: No AI-attributable outcome: the AI-assisted editor was found slower and was disabled. Organizations: No AI-attributable outcome: responders' use of OSM data is documented, but not tied to AI-derived features. The AI tool reduced mapping speed. Society: No AI-attributable outcome documented.
Observed or intended: Not assessed
Rubric: Not assessed
Assessor and date: Not assessed
Alternative assessments: Not assessed
Fill: The deployer's account

Entry details

Kind: Not assessed
Related entries: Not assessed
Field: Humanitarian mapping / disaster response. Volunteer humanitarian mapping.
Purpose: AI-assisted mapping was encouraged to speed building and road mapping over large areas for responders. The evidence shows it slowed mapping and lowered quality in this activation. The claimed benefit was not delivered.
Dates: Date basis unassessed: 2023 (precision unassessed). Updated: 21 October 2026.
Stage: Not assessed
Version: SHA-256: f70f534fb14b…

References

Other identifiers: Not assessed
Category mappings: Not assessed
Brief: Not assessed
Revisions: 4 revisions. Latest: 21 October 2026.

Sources

Report Deployer

Humanitarian OpenStreetMap Team (2023). Turkey and Syria Earthquake Response February 2023 Size-up. hotosm.org, Internet Archive capture of 29 May 2023.
Press release Deployer

Humanitarian OpenStreetMap Team (2023). Türkiye Earthquakes, February 2023 Activation. hotosm.org disaster services page, Internet Archive capture of 7 February 2023.
Other Deployer

OpenStreetMap wiki (2023-2026). 2023 Turkey Earthquakes, section 'Be wary of AI assisted mapping (RapiD)'. Community wiki page, last modified 5 June 2026.