LLMMedium
Startup Misled Research Participants about GPT-3 Use in Mental Healthcare Support
- Occurred
- Dec 1, 2022
- Published
- Jan 10, 2023
- Source
- AI Incident Database
Summary
OpenAI's GPT-3 was deployed by a mental health startup without ethical review to support peer-to-peer mental healthcare, and whose interactions with the help providers were "deceiving" for research participants.
Developers
- OpenAI
Deployers
- Koko
Harmed parties
- research participants
- Koko customers
Classification
Category confidence: 90%
Severity confidence: 85%
Rationale: GPT-3 misuse in mental healthcare without ethical review constitutes a notable failure with potential harm.
Classifier: Qwen/Qwen3.5-35B-A3B-FP8