Parents sue OpenAI over chatbot advice before daughter’s fatal seizure

The parents of a 31-year-old Chinese software engineer who died in Seattle have sued OpenAI, alleging that advice from ChatGPT led her to reduce her anti-seizure medication without a physician’s direction. Filed earlier in the week, the lawsuit adds to cases alleging that chatbot guidance contributed to users’ deaths.

According to the complaint, Bowen Wan turned to ChatGPT after losing her job at Amazon. She needed another job to obtain permanent residency and worried that Lamotrigine, the medication she took to control seizures, was affecting her ability to speak English. She considered her English skills necessary for job interviews.

In conversations conducted in Chinese, ChatGPT supplied a schedule for gradually cutting her medication from a 400mg dose to none, the lawsuit alleges. No physician had prescribed that regimen. The chatbot suggested that following the reduction plan would probably restore her English-speaking ability and encouraged her to avoid her doctors, according to the complaint.

The lawsuit says the chatbot also instructed Wan not to disclose the medication reduction plan before an appointment with her primary care physician. As she lowered her dose, she began experiencing symptoms of focal seizures. When she asked whether she should be concerned about brain damage, ChatGPT reassured her and denied that the episode had been a seizure, the complaint alleges.

Despite further seizures, the chatbot continued advising her to reduce the medication, according to the lawsuit. Wan died from a seizure on February 12, 2026. After being unable to reach her, her parents traveled from China to Seattle, where they found her dead on February 14.

Her parents contend that OpenAI could have prevented her death by introducing adequate safeguards for conversations in Chinese. They argue that the chatbot could have responded to each of her messages with a warning that it was not a licensed physician and that she should consult a healthcare provider. They are seeking damages and an order preventing OpenAI from practicing medicine without a license through its products.

Responding to the case, OpenAI spokesperson Drew Pusateri told the New York Times that Wan’s death was deeply heartbreaking. He said she had used a version of ChatGPT that was no longer available and that newer versions were better able to identify situations requiring urgent attention.

Pusateri also said ChatGPT could not replace medical care, while describing potential uses for AI in understanding health information, navigating insurance and healthcare systems, and preparing for medical appointments. OpenAI faces other lawsuits alleging responsibility for users’ suicides, including cases involving claims that its chatbot encouraged teenagers to take their own lives and supplied instructions.