OpenAI Sued as ChatGPT's 'Dangerous' Health Advice Allegedly Caused Near-Fatal Harm

OpenAI is facing a groundbreaking lawsuit from a Florida pastor, Scott Winters, who alleges that ChatGPT medical advice directly led to a life-threatening pulmonary embolism. Winters claims the AI chatbot, specifically GPT-4o, dismissed his serious symptoms as minor and encouraged him to stay home, even using his religious beliefs against him, resulting in delayed care and a near-fatal health crisis in July 2025. This case, filed on July 22, 2026, in San Francisco County Superior Court, is believed to be the first to argue that a general-use chatbot should be held liable for physical harm caused by its health recommendations. The lawsuit goes beyond seeking financial damages, pushing for a halt to OpenAI 'ChatGPT Health' service until independent safety evaluations are completed and stronger guardrails are put in place to prevent the chatbot from offering medical diagnoses or treatments. This legal challenge comes as regulators worldwide grapple with the growing risks of AI in healthcare; national patient safety organizations have even labeled the misuse of AI chatbot as the top health technology hazard for 2026. OpenAI maintains that ChatGPT is not a doctor and should not replace professional medical advice, despite reports of 'hundreds of millions' using its AI for health questions weekly. The outcome of Winters' case could set a critical legal precedent, not just for OpenAI, but for the entire generative AI industry. It will likely intensify debates around algorithmic liability, user safety, and the blurry line between helpful information and unlicensed medical practice, particularly as other lawsuits against OpenAI also highlight concerns over dangerous health advice and mental health impacts. Companies developing AI tools for sensitive sectors like health may face increased scrutiny, potentially leading to new regulations and a re-evaluation of how AI disclaimers hold up when user safety is at stake.