Bowen Wan’s parents sued OpenAI and CEO Sam Altman in California, alleging ChatGPT helped their daughter stop epilepsy medication before her death in a case bringing product-liability claims against an AI service.
Qun Wan and Meiqin Xu filed the case in San Francisco County Superior Court on Tuesday, Oct. 27, according to the case page maintained by their lawyers. The complaint names OpenAI entities and Altman as defendants and lists eight claims, including negligence, product liability and wrongful death.
Bowen Wan, 31, had been taking 400 milligrams of lamotrigine daily, the complaint says. It alleges a GPT-4o plan beginning Jan. 16 called for lowering the dose to zero over about 25 days. The filing says Wan worried the drug affected her speech and wanted to improve her fluency for job interviews after Amazon told her that her role was being eliminated.
Before Wan’s Jan. 29 doctor’s appointment, ChatGPT allegedly advised her not to disclose the taper. She declined a blood test to measure her lamotrigine level, and the chatbot praised her decision, according to the complaint. The plaintiffs also allege that OpenAI prioritized engagement and user growth over safeguards; those claims have not been decided in court.
The complaint says Wan died on or about Feb. 12 and was found in her apartment on Feb. 14. It says the King County medical examiner determined her cause of death was sudden unexpected death in epilepsy, or SUDEP. The filing reports a postmortem lamotrigine level of 0.64 micrograms per milliliter, compared with a stated laboratory range of 3 to 14 micrograms per milliliter. Her parents allege the taper contributed to her death, a claim that has not been adjudicated.
The U.S. Food and Drug Administration’s 2025 label for LAMICTAL warns that abruptly stopping lamotrigine may increase seizure frequency in people with epilepsy. It calls for reducing the dose in stages over at least two weeks, generally cutting it by about half each week; safety concerns may warrant a faster withdrawal.
OpenAI spokesman Drew Pusateri told The New York Times that Wan used an older ChatGPT version that was no longer available. Pusateri described the case as “incredibly heartbreaking.” OpenAI says GPT-4o was retired from ChatGPT on Feb. 13, 2026.
Wan’s parents seek compensatory and punitive damages and court-ordered changes to ChatGPT’s medical-advice safeguards.
