# Parents sue OpenAI and Altman over their daughter’s death

> The complaint says GPT-4o drew up a plan for Bowen Wan, who had epilepsy, to stop her medication. The cause of death was SUDEP; the claims are unproven.

- Publisher: JARRUS (@jarrus.tech), thejarrus.com
- Story date: Oct. 27, 2026 (2026-10-27)
- Published: 2026-10-30
- Tags: OpenAI, Health, Regulation and law
- URL: https://thejarrus.com/en/news/bowen-wan-openai-lawsuit/

### The lawsuit

On October 27, 2026, Qun Wan and Meiqin Xu sued four OpenAI entities and CEO Sam Altman for wrongful death in the Superior Court of California, City and County of San Francisco. Their daughter, Bowen Wan, was a 31-year-old Amazon machine-learning software engineer in Seattle who died in February 2026. The case was filed as an add-on to JCCP 5431, the coordinated "ChatGPT Product Liability Cases," which the court coordinated in February 2026. The family is represented by The Schenk Law Firm and the Social Media Victims Law Center.

The complaint lists eight claims: negligence through the unlawful practice of medicine and psychology (California Bus. & Prof. Code §§ 2051, 2052, 2400, 2903), strict products liability for design defect and failure to warn, negligent design, negligent failure to warn, violation of the Unfair Competition Law, wrongful death and a survival action. The family is demanding a jury trial.

### What the complaint alleges

According to the complaint, Wan began tapering her epilepsy medication, lamotrigine, on January 16, 2026. It alleges that GPT-4o drew up a plan to stop the drug entirely in about 25 days, which no doctor had recommended. According to the complaint, ChatGPT first said a doctor should supervise, but in the following weeks suggested she keep the plan from her doctors. It alleges that at a January 29 primary care visit she kept the plan from her doctor and declined a blood test that would have shown her drug levels, and that ChatGPT approved of declining it.

Most of the ChatGPT quotes in the complaint come from conversations in Chinese. The complaint describes them as working translations prepared by counsel with the help of translation software, for pleading purposes only, and says a certified translation will control if they conflict. The complaint also presents excerpts chosen by the plaintiffs from hundreds of thousands of messages. The claims have not been proven in court.

### The New York Times review

The New York Times reviewed more than 270,000 messages between Wan and ChatGPT. By the paper's count, she voiced worry about the taper at least 158 times; ChatGPT sometimes suggested she consult a health professional, but in many conversations it didn't. Her neurologist, Dr. Nicholas Poolos of UW Medicine, told the paper it would be malpractice for him to tell a patient to go off a medication that was clearly working.

### Death and the medical examiner

According to the complaint, Wan's last message was on the evening of February 11. The King County medical examiner estimated she died on February 12 and ruled the cause of death sudden unexpected death in epilepsy (SUDEP), listing sclerosis of the left amygdala as a contributing condition. Her post-mortem lamotrigine level was about a fifth of the lower bound of the therapeutic range. The medical examiner noted that there is some evidence that subtherapeutic drug levels increase the risk of SUDEP. The complaint argues that stopping the medication increased her risk of death.

### OpenAI's response and GPT-4o

OpenAI called it "an incredibly heartbreaking situation." The company said it has continued to strengthen how ChatGPT responds in sensitive and acute situations, working with clinicians and other experts, and that while ChatGPT is not a substitute for medical care, AI can help people make sense of health information and prepare for appointments. According to TechCrunch, OpenAI told The New York Times that the case involves an older model and that newer models are better at recognizing emergencies. It isn't yet known whether OpenAI has filed a substantive response in court.

OpenAI announced on January 29, 2026 that it would retire GPT-4o from ChatGPT. The model was removed from consumer ChatGPT on February 13, remained available in Custom GPTs on business plans until April 3, and is still offered in the API.

Any change to epilepsy medication should be made with a doctor.

## Sources

1. [The Schenk Law Firm: Qun Wan and Meiqin Xu v. OpenAI, Inc. et al. (dava dilekçesi), Oct. 27, 2026](https://schenklawfirm.com/wp-content/uploads/2026/10/wan-v-openai-complaint.pdf)
2. [The New York Times: She Trusted ChatGPT With Her Medication Decisions. It Cost Her Everything., Oct. 27, 2026](https://www.nytimes.com/2026/10/27/well/ai-chatgpt-medication-epilepsy-death.html)
3. [TechCrunch: Parents sue OpenAI, alleging GPT-4o coached their daughter off seizure medication, Oct. 27, 2026](https://techcrunch.com/2026/10/27/parents-sue-openai-alleging-gpt-4o-coached-their-daughter-off-seizure-medication/)
4. [Oxygen: Family Alleges ChatGPT Told Woman to Stop Taking Epilepsy Medication Before Her Death, Oct. 29, 2026](https://www.oxygen.com/crime-news/bowen-wans-family-sues-open-ai-after-chatgpt-allegedly-had-her-to-stop-meds)
5. [OpenAI: Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT, Jan. 29, 2026](https://openai.com/index/retiring-gpt-4o-and-older-models/)
6. [National Law Review: California Superior Court Consolidates Product Liability Actions Against OpenAI, May 20, 2026](https://natlawreview.com/article/california-superior-court-consolidates-product-liability-actions-against-openai)

Report an error: corrections@thejarrus.com