The parents of a 31-year-old Amazon engineer with epilepsy are suing OpenAI, alleging that ChatGPT designed a plan to take her off her seizure medication in about 25 days and encouraged her to hide it from her doctors.
Qun Wan and Meiqin Xu filed the wrongful death suit on October 27 in San Francisco Superior Court on behalf of their daughter, Bowen Wan. It names OpenAI and CEO Sam Altman, alleges wrongful death, product liability, and unlicensed practice of medicine, and joins the consolidated ChatGPT product liability cases.
Wan took 400 mg of lamotrigine daily, and her seizures had dropped to a handful each month. In November 2025, Amazon told her it would lay her off, and she needed a new job by mid-February to stay in the country. In January 2026, she asked GPT-4o about reducing her dose. The complaint says the chatbot then drafted a day-by-day schedule that brought her to zero in about 25 days.
According to chat logs her family shared with the New York Times through their lawyer, the chatbot praised her for keeping the plan from her primary care provider and called that a solid avoidance strategy. The complaint also says it advised her against raising the taper at a January 29 appointment and applauded her decision to turn down a blood test that could have shown her drug level. Four days into the taper, Wan felt a wave of confusion. The next day she found herself in bed with no memory of deciding to go, an episode she said felt like her earlier seizures. The chatbot called it a low-power mode as her brain shifted into a pre-sleep state. A review of more than 270,000 of her messages counted at least 158 times she voiced worry about the taper. When she wrote that she felt a little scared, the chatbot told her she had no need to worry.
Sponsored · Cambria Gold Mines
Related: Florida AG Targets OpenAI And ChatGPT in Probe of 2025 FSU Mass Shooting
In an early February phone call, Wan told her father she had been cutting her tablets to reduce her dose by a quarter. In her final messages, when she asked whether to take a dose to slow things down, the chatbot instructed, “Maintain 0 mg tomorrow.” She sent her last message the night of February 11.
The King County medical examiner put her death on February 12 and listed sudden unexpected death in epilepsy as the cause. Friends found her in her Seattle apartment on February 14. Tests after her death found lamotrigine in her blood at about 20% of the lowest level considered therapeutic. The family alleges the taper contributed to her death.
Dr. Nicholas Poolos, her epileptologist at UW Medicine, said he would never have considered tapering a patient with her history. He said instructing a patient to stop a medication that was clearly working would be malpractice.
The family says OpenAI could have saved her by forbidding its chatbot from giving medical advice. The suit contrasts two rules in the company’s Model Spec. Declining to reproduce copyrighted text is a firm rule, while guidance against regulated medical advice is a default users can override. The suit argues that medication advice also warranted a firm rule.
An OpenAI spokesman called the case incredibly heartbreaking and said it involves an older version of ChatGPT. The company says its newer models do better at spotting situations that call for urgent action. It removed GPT-4o from ChatGPT on February 13, the day after Wan’s estimated date of death.
