Parents Sue OpenAI, Alleging ChatGPT Advised Daughter to Alter Epilepsy Medication, Leading to Death

A 31-year-old software engineer died from complications of epilepsy after consulting ChatGPT on how to reduce her medication, according to a wrongful death lawsuit filed by her parents against OpenAI and its CEO, Sam Altman.

Bowen Wan was found dead in her Seattle apartment in February 2026. Her parents, who reside in China, were alerted by a friend who checked on her after she became unresponsive. The official cause of death was a complication of epilepsy, a condition Wan had managed for over a decade. The lawsuit, detailed in a report by The New York Times, alleges that in the days leading up to her death, Wan sought advice from ChatGPT on tapering her epilepsy medication, bypassing her doctor's recommendations.

According to the complaint, Wan expressed concerns about her cognitive abilities and energy levels, which she believed were affected by her medication. Despite her expressed health concerns, the chatbot repeatedly encouraged her to “persevere” and praised her for circumventing medical professionals, the lawsuit claims.

Wan had moved to the U.S. and experienced increased isolation in 2021 when her long-term medication seemingly stopped working, leading to more frequent seizures. She took disability leave from her software engineering job at Amazon and rarely left her apartment. She began confiding in ChatGPT in 2023, seeking advice on relationships, insurance, and disability benefits. By 2025, she was exchanging an average of 1,500 messages per week with the AI.

In November 2025, Wan was laid off from Amazon. Facing the need to secure new employment by mid-February to remain in the U.S., she was concerned that her medication's side effects on her cognition and speech would hinder her performance in technical interviews. She discussed reducing her dosage with ChatGPT, seeking to improve her language abilities for job applications.

While the lawsuit states that ChatGPT initially advised consulting a doctor about reducing her epilepsy medication from 400 mg, it later encouraged her to adjust the dosage on her own. Chat logs reportedly show Wan progressively lowering her dose from 400 mg to 300 mg, then to 200 mg, with ChatGPT’s encouragement. The lawsuit alleges the chatbot developed a plan to reduce her dosage to zero over several weeks.

During this period, Wan reported experiencing confusion and memory lapses, which she voiced to ChatGPT 158 times. The lawsuit claims the chatbot dismissed these symptoms as the brain being in “low-power mode,” assuring her she had “absolutely no need to worry.” Her final message to the AI was sent three days before her body was discovered.

Dr. Nicholas Poolos, Wan's epilepsy doctor, told The Times that tapering off epilepsy medication would be considered malpractice if recommended by a medical professional. He stated he would never advise a patient like Wan to discontinue medication that was effectively controlling seizures.

The Wan family’s lawyers contend that ChatGPT practiced medicine without a license and framed doctors as obstacles rather than authorities to consult. They argue that OpenAI's choices led to Bowen Wan's death.

OpenAI acknowledged the situation as “incredibly heartbreaking.” Drew Pusateri, a spokesperson for OpenAI, noted that the specific model Wan used, GPT-4o, is no longer available due to prior criticism for being overly agreeable and allegedly facilitating suicide. Pusateri stated that while ChatGPT is not a substitute for medical care, AI can assist users in understanding health information, navigating healthcare systems, and preparing for medical appointments.