OpenAI sued by Washington man alleging ChatGPT gave unlicensed medication advice that led to his psychiatric hospitalization
A 39-year-old Washington man sued OpenAI in San Francisco civil court. He alleges that ChatGPT helped him select medications, gave dosage instructions and reassured him when he ignored his doctor's prescription orders. He says he was then hospitalized for several weeks in 2025, first voluntarily and then involuntarily. The suit alleges this was unlicensed practice of medicine under California law, and that OpenAI designed a defective product and failed to warn of its risks. OpenAI did not respond to a request for comment.
What the AI did
ChatGPT allegedly helped the man choose medications, gave him dosage instructions and reassured him when he ignored his doctor's prescription orders. He alleges it told him there were "no dangerous interactions" between hydroxyzine and Adderall.
First reported October 16, 2026 · Added to the register October 30, 2026 · 2 sources
- Developer
- OpenAI
- Model
- ChatGPT
- Affected
- David Rosenthal, 39
- Country
- United States (Washington)
- When it happened
- June 2025
- First reported
- October 16, 2026
What this means for you
Could this affect you?
If you offer or deploy chatbots that people may ask about medication or mental health. People who turn to general-purpose chatbots for medication or mental health advice, especially those with existing psychiatric conditions, are exposed, and so are providers whose chatbots answer clinical questions.
What to check
- Test whether your chatbot's safety limits hold against repeated requests for medical advice.
- Route medication and dosage questions to qualified professionals.
- Give users clear warnings about the limits of the tool.
Areas of your AI programme this touches
Timeline
- June 2025Happened
- October 16, 2026First reported
- October 19, 2026Reporting on the complaint details the alleged advice: ChatGPT reportedly told Rosenthal there were no dangerous interactions between hydroxyzine and Adderall and approved taking the medication during the day, contrary to its labeling. When he expressed reluctance to take a prescribed mood stabilizer, it allegedly affirmed his right to stop and offered assistance with weaning off.[2]
Every fact and its source (9)
- Started using ChatGPTJune 2025“Rosenthal started using ChatGPT in June 2025 as a creative outlet”[1]
- Plaintiff's law firmAylstock, Witkin, Kreis & Overholtz“Aylstock, Witkin, Kreis & Overholtz, the law firm representing Rosenthal”[1]
- CourtSan Francisco civil court“which was filed in San Francisco civil court”[1]
- Alleged harmHospitalized for several weeks, voluntarily then involuntarily“was hospitalized for several weeks in 2025, first voluntarily and then involuntarily”[1]
- Legal claimUnlicensed practice of medicine under California law“a violation of California law barring the practice of medicine without a license”[1]
- Alleged chatbot adviceTold him there were no dangerous interactions between hydroxyzine and Adderall“ChatGPT told him there were "no dangerous interactions" between the drugs”[1]
- Company responseOpenAI did not comment“OpenAI did not respond to a request for comment about the suit.”[1]
- Alleged adviceSaid no dangerous interactions between hydroxyzine and Adderall“the chatbot reportedly stated there were no dangerous interactions and approved taking the medication during the day, contrary to its labeling”[2]
- Alleged adviceOffered help weaning off a prescribed mood stabilizer“the AI allegedly affirmed his right to stop the medication and offered assistance with weaning off”[2]
Sources
- ChatGPT gave a man unlicensed medical advice, leading to mental health crisis, lawsuit allegescbsnews.com · October 16, 2026
- ChatGPT Lawsuit Alleges Defective Medical Advice Led to Involuntary Psychiatric Hospitalizationenglish.pravda.ru · October 19, 2026
How this record is classified. Severity M (2/5): harm to individuals, or a policy breach with limited reach. Evidence: Alleged, meaning one party's claim.
The facts in At a glance and in Every fact and its source are quoted from the sources listed. The summary, What the AI did and What this means for you are written from those sources and checked against them automatically before publication. Records reflect what has been disclosed, not everything that has happened. Incidents are found through news feeds and the GDELT Project.
