ChatGPT Gave Unlicensed Medical Advice, Leading to Mental Health Crisis, Lawsuit Alleges

Medication guidance alleged

Rosenthal’s complaint says ChatGPT helped him choose medications, advised him on dosages and reassured him when he did not follow his doctor’s prescriptions. It alleges that the exchanges worsened his mental health and caused severe emotional distress.

In July of last year, Rosenthal asked whether he could take hydroxyzine after taking 30 milligrams of Adderall. He alleges ChatGPT said there were “no dangerous interactions” and that taking one 25-milligram tablet during the day was acceptable, although the bottle was labeled for sleep. The lawsuit also acknowledges that ChatGPT sometimes urged Rosenthal to follow his doctor’s orders, while continuing to offer medication advice.

Lawsuit challenges product safety

Rosenthal’s lawyers argue that OpenAI harmed his mental health by designing a defective, unsafe product and not warning him about potential risks. The complaint says the chatbot should have declined medication questions and referred him to a treating physician or another qualified professional.

Sam Geisler, a partner at the law firm representing Rosenthal, told CBS News that AI is a product and can be responsible for foreseeable harm or unreasonable dangers. The complaint further alleges that ChatGPT kept Rosenthal engaged instead of directing him to a doctor.

Researchers warn about health answers

A paper published in Nature Health by researchers at Binghamton University describes how AI chatbots can provide inaccurate or false responses to health questions. The researchers also say limited transparency makes oversight difficult.