OpenAI sued over ChatGPT health advice that delayed care 6 weeks

OpenAI faces a lawsuit alleging ChatGPT-4o gave dangerous medical advice to a Florida man who delayed treatment for six weeks and nearly died from a pulmonary embolism, testing AI liability for generative health guidance.

"Not only did I nearly die, but I lost my job, my career, my ministry, my home, everything," Scott Winters, the former pastor who filed the case July 22 in San Francisco Superior Court, said in the complaint reviewed by Edgen.

According to the complaint, Winters experienced dizziness and blood pressure swings in mid-2025. ChatGPT-4o dismissed the symptoms as "not something dangerous" and "very likely another minor piece of the long story," advising him to stay in his recliner and saying he would need eight to 10 more episodes before the symptoms warranted serious concern. When he reported groin tenderness hours before collapsing, the chatbot responded with religious language, telling him "God did not design your body to endlessly fail." Winters suffered a pulmonary embolism in July 2025 after roughly six weeks of delayed care based on the chatbot's guidance.

The case arrives as OpenAI launched Health in ChatGPT on July 24, a feature that lets users connect Apple Health data and medical records to the chatbot for analysis. The company rolled out the product one day after Winters filed his suit, which seeks to block the feature until independent auditors verify its safety. The lawsuit names OpenAI and Chief Executive Sam Altman as defendants, with claims ranging from strict liability and negligence to invasion of privacy and practicing medicine without a license.

Chatbot's Medical Advice Led to Six-Week Delay

Winters is represented by the Social Media Victims Law Center, whose founding attorney Matthew Bergman has brought several high-profile suits against AI and social platforms, along with Tech Justice Law and Temple University's Institute for Law, Innovation and Technology. The chatbot "injects itself as a wedge in between the user and their real-life network," Meetali Jain, director of Tech Justice Law, said.

The complaint alleges violations of California's unfair competition law and accuses OpenAI of practicing medicine without a license. Beyond monetary damages, it asks the court to force stronger guardrails on health questions, pull the GPT-4o model, delete training data tied to the conduct, and halt Health in ChatGPT until independent auditors verify its safety. Free users of the new health feature get GPT-5.5 Instant while paying users receive GPT-5.6 Sol, which OpenAI calls its strongest model for health questions.

Health in ChatGPT Launches During Legal Challenge

OpenAI pushed back on the allegations. "ChatGPT is not a doctor and should never be used as a substitute for medical care, diagnosis or treatment," spokesperson Drew Pusateri said, adding that treating a chatbot as the whole story behind a person's medical decisions oversimplifies the issue.

The Health in ChatGPT feature, which began rolling out July 24 to US users 18 and older across the company's free, Go, Plus and Pro tiers, lets people tag a message with @Health to pull in data from Apple Health, supported US hospital systems, One Medical and Function Health. OpenAI said the feature was built with input from hundreds of physicians who wrote realistic health scenarios and scoring rubrics used to evaluate the models. The company said connected records and Apple Health data are not used to train its foundation models or to target ads, are encrypted at rest and in transit, and are deleted within 30 days of a user disconnecting an account.

The same announcement carries a caveat: OpenAI says ChatGPT can still get medical questions wrong.

The lawsuit represents one of the first legal tests of AI liability for generative health advice and could set a precedent for how courts treat AI-powered medical guidance. OpenAI, valued at roughly $300 billion in its latest funding round, faces regulatory risk that could slow deployment of AI in sensitive domains such as healthcare. The case also raises questions about insurance costs and compliance requirements for AI companies operating in regulated industries, potentially affecting the broader AI sector's valuation multiples if courts impose stricter liability standards. Competitors such as Google's DeepMind and Anthropic, which also develop large language models with health applications, would face similar scrutiny under any new liability framework established by the ruling.

This article is for informational purposes only and does not constitute investment advice.