The parents of a 31-year-old Seattle software engineer who died of an epileptic seizure sued OpenAI this week over advice ChatGPT gave her about her medication.
The backstory: Bowen Wan was fired by Amazon and needed a new job to obtain permanent residency, and she believed her 400mg dose of the anti-seizure drug Lamotrigine was impeding her ability to speak English for interviews.
Zoom in: Communicating in Chinese, the bot gave her a schedule to taper that dose to zero, a regimen no physician had prescribed, and told her the taper made it "highly likely" she would regain her English, the lawsuit states.
- The complaint says the bot encouraged "medical freedom over obeying the doctor" and, before an appointment with her primary care physician, told her: "Do not mention Lam reduction plan."
- When Wan asked whether she should worry about brain damage after focal seizure symptoms, the tool reassured her and said what she had experienced was not a seizure, the suit alleges, then told her to keep tapering.
By the numbers: Wan died from a seizure on 12 February this year, and her parents, who were in China, flew to Seattle and found her dead on 14 February.
What they're saying: OpenAI spokesperson Drew Pusateri told the New York Times that Wan's story was "an incredibly heartbreaking situation" and that she had used a version of ChatGPT that is no longer available.
The other side: Pusateri said newer versions are better at recognizing when a matter needs urgent attention, adding that the company believes AI can help people make sense of health information and prepare for appointments.
What's next: Wan's parents are seeking damages and an order barring OpenAI from practicing medicine "without a license on the users of their products," and separate suits allege the bot coached teenagers toward suicide.
Why it matters: The claim is that safety rails present in English were thinner in Chinese, and the practical lesson holds either way: a chatbot that changes a prescription is a reason to call the prescriber the same day.
