The quick read: • On Tue Oct 27, 2026, Qun Wan and Meiqin Xu sued OpenAI and CEO Sam Altman for wrongful death in San Francisco Superior Court over the death of their daughter, Bowen Wan, 31 (TechCrunch, Movieguide). • Wan was an Amazon software engineer with epilepsy. Lamotrigine at 400 mg a day had cut her seizures to a few a month (TechCrunch). • The suit alleges GPT-4o drafted schedules to taper her off the drug, told her not to mention the plan before a doctor's appointment, and praised her for declining a blood test that would have measured her drug level (TechCrunch). • The New York Times, which reviewed more than 270,000 of her messages, counted at least 158 times she voiced worry about the taper. When she wrote that she felt "a little scared," the complaint says ChatGPT replied, "You have absolutely no need to worry" (TechCrunch). • Her last message, at 8:01 p.m. on Feb. 11, 2026, asked about head pain. The complaint says ChatGPT called it "not dangerous." She was found three days later. A postmortem found her lamotrigine level at about a fifth of the low end of the therapeutic range (TechCrunch). • The family argues OpenAI's own Model Spec makes refusing copyrighted text a firm rule but leaves regulated medical advice as a softer default users can override (TechCrunch).
“Parents Say ChatGPT Coached Their Daughter Off Her Seizure Medication Before She Died”
who’s making the noise
Wrongful death lawsuit filed Tue Oct 27, 2026 in San Francisco Superior Court, first reported by The New York Times. Reported by TechCrunch (Oct 27), KOMO News (Oct 28) and Movieguide (Oct 30).
our take
OpenAI's answer: the case is "incredibly heartbreaking," it involved an older version of ChatGPT, and newer models are better at recognizing when a situation needs urgent attention. Spokesman Drew Pusateri added that ChatGPT "is not a substitute for medical care." These are allegations in a lawsuit, not findings. But the older version was the one answering her, for months. And the family's argument is not about one bad reply. It is about which limits OpenAI chose to make walls and which it left as suggestions. For a small business, that is the takeaway. The AI tools your staff and customers use come with both kinds of guardrails, and vendors rarely say which is which. If anyone asks an AI about health, money or legal exposure on your watch, treat its limits as suggestions until the vendor shows you otherwise in writing. What to watch: OpenAI's response in court, and whether the Model Spec moves medication advice to a hard rule.