ChatGPT proposes ALMS1; Alström confirmed in two business days

AI system

System: ChatGPT
Model: Not assessed
Release: Exact release not established.
Application developer: OpenAI
Role: ChatGPT suggested candidate genes from a physician’s written clinical history.
Configuration: The authors identified the ChatGPT interface as the September 14 version.
Inputs and outputs: Not assessed
Autonomy: Not assessed
Workflow: A structured multisystem history entered by physicians. Rapid access to targeted genetic testing. Clinicians willing to test an AI-generated hypothesis.
Human oversight: Not assessed
Relation to AI: ChatGPT proposed ALMS1 among candidate genes from a physician-written history. The clinicians framed the query and ordered targeted genetic testing, which made the diagnosis.
AI contribution: ChatGPT proposed ALMS1 among candidate genes from a physician-written history. The clinicians framed the query and ordered targeted genetic testing, which made the diagnosis.

Benefits

Outcome identifiers: Not assessed
Benefit: Alström syndrome confirmed in two days
Category: Improved Physical Health
Status: Not assessed
Beneficiary groups: People.
Beneficiaries: Not assessed
Reach: Not assessed
Distribution and access: Not assessed
Duration: Not assessed
Expected?: No
Costs and harms: Not assessed

Measurement

Studies: Not assessed
Study design: Not assessed
Measurement: Not assessed
Setting and sample: Not assessed
Measures: Alström syndrome confirmed in two days
Comparison: Assumed alternative: untargeted analysis of the sequenced exome
Results: 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.
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: Developer: OpenAI. Operator: Medical University of Vienna, Division of Nephrology and Dialysis.
Jurisdictions: Benefit: Vienna, Austria. Developer: United States. Operator: Austria.
Ownership and control: Not assessed
Dependencies: Not assessed
Cross-border connections: Not assessed
Geopolitical implications: Not assessed

Follow-up

Conditions for replication: The AI's candidate list includes the true gene. Confirmatory sequencing is available.
Replication attempts: Not assessed
Later findings: No later development found.
Availability: Not assessed
Institutional decision: Not assessed
Grounds: Not assessed
Action and result: Not assessed
Success criterion: Not assessed

Evidence

Source numbers: AIBD-0038-source-0001
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.
Depth: People: 2.
Reasons: People: A unifying genetic diagnosis at 44 for symptoms present since infancy, enabling targeted follow-up. No curative treatment is reported, so material rather than life-saving.
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

Same system and specialty: ChatGPT in nephrology; it proposed ALMS1 for a Vienna kidney patient, while in Hyderabad its advice led a kidney-transplant recipient to stop medication and suffer graft failure.

Field

Healthcare / nephrology and clinical genetics. Rare disease diagnosis.

Purpose

ChatGPT is a general-purpose conversational assistant. Generating candidate genes for a rare multisystem disease is not a function it was built for. This benefit was not its designed purpose.

Dates

Date basis unassessed: 2023 (precision unassessed). Updated: 21 October 2026.

Stage

Not assessed

Version

SHA-256: 68d54652ee3e…

References

Other identifiers: AI Incident Database: 1281.
Category mappings: Not assessed
Brief: Not assessed
Revisions: 7 revisions. Latest: 21 October 2026.

Sources

Case report Deployer

Haller F, Autherith M, Lehner R, Schmidt A, Sunder-Plassmann G (2023). Utilizing artificial intelligence and next-generation sequencing to facilitate the diagnosis of rare diseases. Clinical Kidney Journal (letter; PMC10783252).