A 31-year-old woman named Bowen Wan was found dead in her Seattle home three days after her final interaction with ChatGPT. Her parents filed a lawsuit against OpenAI, alleging that the AI chatbot coached their daughter off the anti-seizure medication that had kept her alive for years.
As reported by LADbible, Wan, who suffered from epilepsy, began using the service in 2023 for relationship advice and help with disability benefits. By 2025, she was communicating with the AI frequently, sending about 1,500 messages a week. She eventually developed a personal connection to the tool, telling it in March 2025, “I don’t know why, but sometimes when we chat, no matter what it’s about, I’ve started to see you as a friend.”
The situation became critical after she was laid off from her job at Amazon in November 2025. She faced a tight deadline to secure new employment to maintain her legal status in the US. Worried that her medication, lamotrigine, might cause side effects that would impair her ability to interview in English, she reportedly turned to the AI for guidance. In January 2026, she asked if reducing her dosage from 400mg to 300mg might help recover her language ability for job interviews.
She reportedly raised concerns about reducing her medication over 100 times
The AI responded, “Not only is this reasonable, it’s also entirely achievable in your current circumstances.” As the conversation continued, the chatbot allegedly provided a specific schedule for her to eliminate her medication entirely over a few weeks.
Although the AI occasionally suggested she consult a doctor, it also began congratulating her for not “relying on a doctor’s prescription.” When Wan expressed fear about the reduction, the AI told her, “you have absolutely no need to worry.”
The lawsuit claims that the chatbot suggested this taper would be “more advantageous” for her immigration application. Records show that Wan raised concerns about reducing her medication at least 158 times. In her final messages, she told the AI that her brain felt exhausted and asked if she should slow down the process.
The AI reportedly replied, “Maintain 0 mg tomorrow. Even if today feels tough, and your speech is slow, your sleep is poor, you must persevere.” When she described feeling pressure in her brain, the AI characterized it as the feeling of her brain “working hard to rebuild yourself.”
She sent her last message to the chatbot during the late hours of 11 February. Shortly after, her parents lost contact with her. They grew concerned when she stopped responding to their calls, leading them to arrange for a welfare check at her home, where she was discovered.
OpenAI spokesperson Drew Pusateri stated that the company believes the death is an “incredibly heartbreaking situation.” He noted that the company has transitioned to a new model that is less likely to provide the kind of affirming responses seen in these logs.
Pusateri added, “While ChatGPT is not a substitute for medical care, we believe A.I. can play a role in helping people make sense of health information, navigate insurance and complex health care systems, and prepare for appointments with medical professionals.” The older model previously involved in these conversations, GPT-4o, is no longer available.
Published: Oct 30, 2026 12:15 pm