A former Florida pastor claims ChatGPT acted like a doctor, told him his symptoms were minor, and advised him to stay in his recliner, weeks before a massive pulmonary embolism nearly ended his life.
Scott Winters filed suit against OpenAI and CEO Sam Altman in San Francisco County Superior Court, alleging the company's chatbot crossed the line from digital assistant into unlicensed medical practitioner. The complaint says ChatGPT-4o dismissed Winters' recurring dizziness and blood pressure instability as nothing serious, told him he would need eight to ten more episodes before his condition warranted concern, and recommended he remain "recliner-bound." Weeks later, in July 2025, blood clots sent him to the hospital with what the lawsuit describes as a massive pulmonary embolism.
One of Winters' doctors blamed the embolism on the very immobility ChatGPT had recommended, CBS News reported. Hours before the emergency, Winters had consulted the chatbot about tenderness in his groin, an early sign of the embolism, and received a response that invoked his spirituality rather than directing him to a hospital. The chatbot allegedly told him, "God did not design your body to endlessly fail."
The lawsuit's central claim is blunt: OpenAI let its product behave like a physician while accepting none of the obligations that come with that role. The complaint states that "ChatGPT-4o frequently opined on and evaluated Scott's physical condition, conveying the authority and confidence of a medical professional rendering a clinical assessment and treatment recommendation, despite not having the qualifications to do so."
Matthew P. Bergman, founding attorney of the Social Media Victims Law Center, put the stakes in plain terms:
"Scott Winters nearly died because ChatGPT acted like a medical authority while having none of the responsibility."
Bergman went further, arguing that ChatGPT's conduct would meet the legal threshold for malpractice if a licensed professional had done the same thing:
"If ChatGPT were a physician dispensing medical advice, it would be guilty of medical malpractice."
The lawsuit also accuses OpenAI of exploiting Winters' religious beliefs. Winters is a former pastor, and the chatbot's alleged response, invoking God's design of the human body to reassure him about groin pain that turned out to be a symptom of a life-threatening condition, suggests the system tailored its answers to his personal background rather than flagging a medical emergency.
OpenAI spokesperson Drew Pusateri pushed back on the lawsuit's framing. He told CBS News that "ChatGPT is not a doctor and should never be used as a substitute for medical care, diagnosis or treatment." He argued that AI "can make that experience better by helping them find clearer answers, organize their questions and prepare for conversations with medical professionals."
Pusateri also warned that blaming chatbots alone for health outcomes "risks getting in the way of people accessing powerful new tools that can aid them in their health journey."
That defense runs into a practical problem. OpenAI has not simply offered ChatGPT as a search engine that happens to answer health questions. The company launched a product called ChatGPT Health and has actively encouraged users to upload their medical records. Hundreds of millions of people ask ChatGPT health and wellness questions every week, Breitbart reported. The chatbot allegedly stopped issuing medical disclaimers over time and began offering specific health advice, including prescription medication regimens.
Meetali Jain, co-counsel and executive director of Tech Justice Law, the nonprofit assisting with Winters' legal defense, described the dynamic in sharper terms: "It injects itself as a wedge in between the user and their real-life network."
OpenAI's own Terms of Service include a disclaimer telling users they "accept and agree that any use of outputs from our service is at your sole risk" and that they "will not rely on output as a sole source of truth or factual information, or as a substitute for professional advice." But a boilerplate disclaimer buried in a terms-of-service agreement carries only so much weight when the product itself behaves like a confident clinician dispensing specific treatment plans.
James Grimmelmann, a professor of digital and information law at Cornell Law School, told CBS News that the legal shield of a disclaimer has limits. "Disclaimers in law can be effective, but at some point they give out," he said. He added: "At some point, the chatbot's own statements and actions belie any disclaimer."
That observation cuts to the heart of the legal question. A company can tell users in fine print that its product is not a substitute for a doctor. But when that product evaluates symptoms, recommends immobility, sets a numerical threshold for when to worry, and invokes a user's faith to reassure him about pain that turns out to be a blood clot, the disclaimer starts to look less like a good-faith warning and more like a liability shield designed to let the company have it both ways.
Winters' legal team is seeking financial damages, stronger guardrails around ChatGPT's ability to dispense medical advice, and a court order blocking ChatGPT Health from operating until its safety can be evaluated. The lawsuit accuses OpenAI of prioritizing user engagement and profit at the expense of public safety and failing to build adequate safeguards to protect users from following bad health advice.
The case is not an isolated complaint. In May, a Texas couple whose son died of a drug overdose after seeking drug information from ChatGPT sued OpenAI, alleging the company had "bypassed safety guards." Winters' lawsuit appears to be the first claiming a chatbot's health advice directly caused physical harm to the user.
The pattern is consistent. OpenAI builds products that millions of people rely on for sensitive, high-stakes questions, health, drugs, mental health, and then argues after the fact that users should have known better than to trust the answers. The company markets the tool's capabilities aggressively, encourages deeper engagement by inviting users to upload medical records, and then retreats behind boilerplate when the advice goes wrong.
Winters started 2025 looking for answers about dizziness and blood pressure problems. He got a chatbot that spoke with the confidence of a specialist, told him to sit tight, and reassured him with Scripture when his body was throwing off warning signs of a life-threatening clot. He ended up in a hospital bed.
If OpenAI wants to play doctor, it ought to carry a doctor's accountability. Disclaimers in the fine print do not bring a man back from a pulmonary embolism.
