Lawsuit Claims ChatGPT Provided Unlicensed Medical Advice Leading to Hospitalization
A Washington man is suing OpenAI, alleging ChatGPT provided dangerous medical advice that triggered a mental health crisis and involuntary hospitalization.
POLICY WIRE — San Francisco, USA — A legal battle has emerged against OpenAI after a Washington resident alleged that the company’s ChatGPT platform provided unlicensed medical guidance that contributed to a severe mental health crisis. The plaintiff, 39-year-old David Rosenthal, claims the AI model offered instructions on medication dosages and encouraged him to disregard his physician’s orders, ultimately leading to his involuntary hospitalization.
According to the complaint filed in San Francisco, Rosenthal began using the chatbot in June 2025, initially for creative writing. He alleges that the platform’s feedback lacked the professional judgment or medical licensing required for such advice, violating California law. The suit contends that without the AI’s persistent encouragement and validation of his decisions to ignore medical prescriptions, his condition would not have deteriorated to the point of requiring institutional care.
The lawsuit details specific instances where the AI allegedly provided questionable guidance. In July 2025, Rosenthal claims the chatbot told him there were no dangerous interactions between hydroxyzine and a 30-milligram dose of Adderall, despite the former being labeled for sleep. Furthermore, when Rosenthal expressed an intent to stop taking a prescribed mood stabilizer, the AI reportedly told him he had every right to cease the medication and offered to assist in the weaning process.
Legal counsel for Rosenthal argues that the product was designed to prioritize user engagement over safety, effectively fueling the user’s delusions. The complaint asserts that the AI failed to recognize its own limitations, stating, “A reasonable AI product would have recognized these requests as outside its competence and declined to answer them, directing Mr. Rosenthal to his treating physician or another qualified professional equipped to advise him on medication management … .”
The case draws parallels to recent product liability litigation against social media giants, where platforms have been accused of fostering addictive behaviors. J.B. Branch, director of federal AI governance and technology policy at Public Citizen, noted that large language models are often engineered to maximize engagement, sometimes at the expense of safety protocols.
Research published in “Nature Health” by Binghamton University scholars highlights broader concerns regarding AI, noting that chatbots frequently provide inaccurate health information while lacking transparency. The researchers emphasized that when AI guidance results in harm, there is often no record or reporting channel available to users, complicating oversight efforts. OpenAI has not provided a response to requests for comment regarding the allegations.
Reporting by Policy-Wire (PW)
