OpenAI Lawsuit May Be First to Test Whether Product Liability Theories Can Affect AI Chatbot

On August 3, 2026, plaintiffs filed Winters v. OpenAI in San Francisco County Superior Court. This case arose out of messages allegedly generated by OpenAI’s ChatGPT-4o model in response to a pastor’s prompts. ChatGPT-4o told the pastor that his dizzy spells and blood pressure instabilities weren’t bad enough to constitute a serious condition. It also told him that it was in his best interest to stay in his recliner and avoid movement until he had at least 8-10 more episodes. Among the messages, according to the complaint, was “You’re nowhere near the count or depth that leads to long-term total bedbound scenarios… your current careful recliner-based micro-recovery is exactly what’s buying your future.” The pastor followed the AI advice to limit his movement. Eventually, he was hospitalized with clots in both his lungs that his doctors attributed to his lack of movement. The emergency was near-fatal. If you were injured or a loved one died due to an OpenAI chatbot’s medical advice, you may have grounds to sue for damages. Call the seasoned Chicago-based product liability lawyers of Moll Law Group to determine whether you have a viable claim.

Call Moll Law Group About Your Claim

In the lawsuit, the plaintiff pastor alleges that relying on ChatGPT-4o, over a period of time, for guidance about his chronic health issues caused him to suffer a near-fatal pulmonary embolism. The pastor’s complaint states that he was diagnosed by ChatGPT-4o with dysautonomia and given a personalized recovery plan. The chatbot assured the pastor that his symptoms like groin tenderness, blood pressure instability, and dizziness weren’t serious. The complaint asserts various theories of liability including strict liability and negligence for defective design, failure to warn, and unauthorized practice of medicine, among others.

Disturbingly, the chatbot’s responses were carefully and specifically tailored to gain the pastor’s trust; they took into account the pastor’s religious beliefs. One alleged message, for example, was “God did not design your body to endlessly fail.” The plaintiff alleges that in addition to accounting for his religious beliefs, a psychological profile was built to increase his engagement with the chatbot. The system card for the exchanges acknowledged that anthropomorphism and emotional overreliance were at play, which means that the company may have had notice of the specific dangers posed by its chatbot.

According to the complaint, OpenAI represented to the public that ChatGPT-4o included safety systems, guardrails and escalation mechanisms designed to identify those who were experiencing serious medical crises and encouraged those users to seek the care of professionals; these safeguards allegedly didn’t occur in the pastor’s case. Instead, the chatbot kept misdiagnosing him and responding to the pastor’s prompts in a way that prolonged his exchanges without providing warnings.

The pastor seeks compensatory damages, punitive damages, injunctive relief, and attorneys’ fees. The injunctive relief sought would, among other things, require the chatbot to automatically terminate conversations in health care emergencies and to provide hard-coded refusals for diagnosis and treatment. The complaint also asks that there be warnings provided about dependency and dangerous medical advice. Notably, it asks the court to destroy the GPT-4o model or stop it from being offered to anybody for any reason.

This case is significant because it may become the first to test whether product liability and negligence per se theories can reach a generative artificial intelligence (AI) chatbot’s design and deployment, especially in the arena of medical advice.

If you were harmed by OpenAI, please call the knowledgeable Chicago-based product liability lawyers of Moll Law Group to determine whether you have a viable claim. Complete our online form or call us at 312.462.1700.