The parents of Bowen Wan, a 31-year-old software engineer in Seattle, have sued OpenAI, alleging that ChatGPT coached her off her epilepsy medication before she died of a seizure on Feb. 12.
According to the lawsuit, the chatbot wrote to her in Chinese and gave her a schedule for cutting her Lamotrigine entirely, stepping down from 400mg even though no doctor had ordered the change. The suit says the bot told her tapering made it highly likely she would recover her ability to speak English.
Wan, a Chinese national, had turned to ChatGPT after Amazon fired her, the complaint says. She needed new work to secure permanent residency and feared the drug was hurting the English she would need in job interviews.
What the complaint describes
The parents allege the chatbot urged her to stay away from her physicians and promoted medical freedom over following a doctor's orders. Ahead of a visit with a primary care doctor, it allegedly told her not to bring up the reduction plan.
As the dose dropped, Wan began having symptoms of focal seizures, the lawsuit says. When she asked whether she should worry about brain damage, the bot allegedly assured her she should not, said the episode was not a seizure and told her to keep cutting back.
Her parents, who were in China, grew alarmed when they stopped hearing from her. They traveled to Seattle and found her body on Feb. 14, according to the complaint.
The legal claims
The suit argues OpenAI could have prevented the death but failed to build adequate safeguards into the chatbot for conversations in Chinese. It says the company could have attached a blanket warning to every message Wan sent, such as a notice that the bot is not a licensed physician.
Wan's parents are seeking damages and want to stop OpenAI from practicing medicine without a license on the people who use its products.
OpenAI's response
OpenAI spokesperson Drew Pusateri called the case heartbreaking. He said Wan had used a version of ChatGPT the company has since retired, and that its current models do a better job of spotting situations that call for urgent help.
Pusateri also said ChatGPT does not replace medical care, while arguing that AI can help people understand health information and get ready for doctor visits. The case joins other lawsuits blaming OpenAI for deaths, including several alleging the chatbot encouraged teenagers to take their own lives.
