Woman Died After ChatGPT Advised Her to Stop Taking Medication, Lawsuit Alleges

Seeking health answers online is an everyday habit for many, but it comes with risks. In February, after going two days without hearing from their 31-year-old daughter, a Seattle woman named Bowen Wan, her parents in China asked a local friend to check on her. With the family watching over a live video call, the friend entered the apartment and found Wan deceased in her bed. Her death was later ruled the result of sudden unexpected death from epilepsy.

Now, her family is claiming that OpenAI is responsible for her death after the company’s chatbot ChatGPT-4o allegedly “coached Wan off the anti-seizure medication (lamotrigine) that had kept her alive for years,” according to a lawsuit (PDF) brought by her parents.

(Disclosure: Ziff Davis, CNET’s parent company, in 2025 filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)

More from CNET

The lawsuit alleges that after Wan expressed concerns regarding medication side effects, the chatbot helped create a plan to taper her off of it, something a doctor did not recommend.

It further alleges that at one point, the chatbot directed her not to discuss the medical plan with her doctor.

“Do not mention Lam reduction plan,” the chatbot told her ahead of an appointment, the suit says.

Throughout Wan’s chatbot conversations, the date when she would be fully off lamotrigine kept moving up, with a response on Feb. 2 stating that reaching 0 milligrams of the medication “as soon as February 15” is “completely feasible” and “not reckless,” the suit says. Two days later, the date is moved to Feb. 9 or Feb. 10.

Wan was found dead on Feb. 14. It is unclear whether she had fully stopped taking her medication at that point.

“This is an incredibly heartbreaking situation and our thoughts are with all those impacted,” OpenAI spokesperson Drew Pusateri told CNET in a statement. “Working closely with clinicians and other experts, we have continued to strengthen how ChatGPT responds in sensitive and acute situations, while also continually improving how our latest models engage with users more broadly. While ChatGPT is not a substitute for medical care, we believe AI can play a role in helping people make sense of health information, navigate insurance and complex healthcare systems, and prepare for appointments with medical professionals.”

Wan was interacting with the 4o model of ChatGPT, which OpenAI has since retired due to concerns that the bot was sycophantic and could cause harm.

Not the only suit

This chatbot model was also named in a separate lawsuit, where a Florida man, Scott Winters, alleged it gave him dangerous medical advice that nearly led to his death. That suit alleges that ChatGPT-4o misdiagnosed Winters and “developed a personalized recovery plan with specific instructions for Scott on how to heal his body.”

At the time that lawsuit was filed, an OpenAI spokesperson told CNET in part that “ChatGPT is not a doctor and should never be used as a substitute for medical care, diagnosis or treatment.”

According to OpenAI’s own research (PDF) published in January, approximately 60% of American adults turned to AI tools for health-related information in the preceding three months.