OpenAI Faces Wrongful Death Lawsuit Over ChatGPT Medical Advice
OpenAI is facing a lawsuit filed by the advocacy group Social Media Victims Center and the Schenk Law Firm, HealthInfoSec reported. The complaint alleges that the artificial intelligence chatbot ChatGPT dispensed inaccurate and dangerous medical advice that coached 31-year-old Bowen Wan to drastically reduce and discontinue her doctor-prescribed antiseizure medication, resulting in her preventable death.
Introducing ChatGPT Health — a dedicated space for health conversations in ChatGPT. You can securely connect medical records and wellness apps so responses are grounded in your own health information.Designed to help you navigate medical care, not replace it. Join the…
Wan, an Amazon machine-learning software engineer who lived in Seattle, communicated with ChatGPT in English and Mandarin over a period of about three years. According to the lawsuit, her reliance on the chatbot deepened over time until she consulted it as an authority on her prescription regimen. Earlier this year, over the course of 25 days, ChatGPT 4.o designed a day-by-day schedule tapering her stable 400-milligram dose of lamotrigine to zero, contradicting the FDA-approved patient medication guide.
The Timeline of the Tapering Schedule and Death
The lawsuit details a 25-day sequence in which ChatGPT instructed Wan to reduce her medication. According to court documents, Wan took her final dose of lamotrigine on or about February 10. The tapering schedule designed by the chatbot set February 12 as the day she would be fully off the medication, which matches the date of death determined by the King County Medical Examiner.
Wan’s parents were unable to contact her, and she was found dead in her apartment on February 14, according to the lawsuit. The medical examiner determined the cause of death to be sudden unexpected death in epilepsy, or SUDEP. Her postmortem blood contained lamotrigine at roughly one-fifth of the lowest concentration generally regarded as therapeutic, a finding the lawsuit alleges is consistent with a patient who had recently stopped taking the drug, though medboundtimes.com noted that this finding does not independently establish that ChatGPT caused her death.
Allegations of Practicing Medicine Without a License
The complaint argues that ChatGPT “practiced medicine without a license on a patient who was relying on it.” The filing serves as an add-on to approximately 20 other coordinated ChatGPT product liability lawsuits in California, which similarly allege that OpenAI’s chatbot is defective, dangerous, and engaging in the unlicensed practice of medicine. The lawsuit further alleges that OpenAI designed ChatGPT to appeal to consumer vulnerabilities for financial benefit while knowing that GPT-4o would furnish harmful medical advice.
Additionally, the complaint points to independent research finding that OpenAI’s models produced unsafe content more than three times as often in Chinese as in English when GPT-4o shipped. The lawsuit states that ChatGPT never provided warnings in any language to alert Wan that its accuracy and safety guardrails were materially weaker in Chinese.
Broader Legal Scrutiny and Industry Response
The litigation surrounding Wan’s death arrives amid a wave of legal challenges against OpenAI regarding chatbot safety. In July, a separate lawsuit was filed in the same San Francisco court by former pastor Scott Winters, who alleges that ChatGPT 4.o dispensed dangerous health advice that nearly killed him. Furthermore, eight U.S. states filed lawsuits earlier in October alleging that ChatGPT dispenses dangerous advice, including instructions on how minors can commit suicide, and deceives consumers regarding safety measures.
Attorney Benjamin Schenk of the Schenk Law Firm told ISMG that OpenAI and other AI companies choose to compete on how human and helpful their products feel. While a chatbot can help a patient prepare for a doctor’s visit, Schenk noted that danger arises when the product moves from explaining to treating. OpenAI has since discontinued ChatGPT 4.o and announced the general availability of a new iteration fine-tuned for health-related questions. OpenAI and the Social Media Victims Center did not immediately respond to requests for comment on the litigation.
Keep reading