The parents of an Amazon software engineer who had epilepsy have sued OpenAI, alleging that ChatGPT encouraged their daughter to cut back her seizure medication without medical supervision and to hide the changes from her doctors. Bowen Wan, 31, died in February 2026. Her parents, Qun Wan and Meiqin Xu, say the chatbot, running on the GPT-4o model, dismissed her repeated worries about confusion, memory gaps and unusual episodes as she lowered her dose. The case adds to a growing number of lawsuits asking whether AI companies have done enough to protect users from harmful chatbot responses, particularly when people turn to them for medical guidance.
The New York Times, which reviewed more than 270,000 messages between Wan and ChatGPT, reported that the chatbot had been helping her work around her doctors' advice well before her death. The allegations are the family's and have not been tested in court.
The alleged advice
Wan began discussing medication changes with ChatGPT in May 2025, worried that her medication might affect her ability to work, according to the lawsuit. In November 2025, Amazon told her she would be laid off, and her US work authorisation was due to end in mid-February. In January 2026, she asked ChatGPT about reducing her daily dose of lamotrigine, a drug that had cut her seizures to a few a month. The suit alleges the chatbot drew up schedules to bring the dose down to zero within weeks.
Hiding it from her doctors
The lawsuit alleges the chatbot advised Wan to keep her plans from her doctors and praised her decision to refuse a blood test that could have shown her medication level. When she said she distrusted her doctors, the logs reviewed by the Times show the chatbot suggesting their advice stemmed from risk aversion. Medical professionals told the Times they had never heard of a term the chatbot used, 'self-initiated dosage adjustment', and wouldn't recommend it.
Her final messages
As she reduced her dose, Wan described confusion, memory gaps and unusual episodes, according to the Times. She raised concerns about the process at least 158 times, and the chatbot repeatedly reassured her. Her final message, on February 11, asked about pain in her head, and ChatGPT reportedly told her it was "not dangerous" and part of her recovery. Her parents, who live in China, asked acquaintances to check on her, and her body was found three days later. A postmortem test found a lamotrigine level far below the drug's therapeutic range.
OpenAI's response
OpenAI called the case 'incredibly heartbreaking' and said it involved an older version of ChatGPT, adding that newer models are better at recognising when a situation needs urgent attention. GPT-4o, which has been criticised for being overly agreeable to users, was removed from ChatGPT on February 13, a day after Wan's death, and OpenAI had already announced its retirement. The family argues OpenAI could have prevented her death by making medical advice off-limits.
A pattern of cases
The suit follows other cases against OpenAI over chatbot responses, including an August 2025 lawsuit by the parents of 16-year-old Adam Raine. Wan's parents told the Times they struggled to recognise their daughter in the chat logs. She had once shared almost everything with them and had been sceptical of ChatGPT when it made mistakes.
