A 55-year-old pastor from Florida experienced life-threatening pulmonary embolisms, but he didn’t seek medical attention until it was almost too late. That’s because he sought guidance from ChatGPT, which evaluated his reported symptoms and dispensed incorrect medical advice.
The man, Scott Winters, together with the organization Tech Justice Law, the Social Media Victims Law Center, and the Institute for Law, Innovation & Technology at Temple University, filed a lawsuit in San Francisco County Superior Court against OpenAI, Inc. (the company that creates ChatGPT) and its CEO Sam Altman. The lawsuit alleges that ChatGPT crossed the line from giving advice to practicing medicine without a license, and it attempts to hold artificial intelligence accountable.
ChatGPT Said to Treat Recurring Dizziness With a ‘Recliner-Based Micro-Recovery’
Scott Winters is a deeply religious evangelical pastor and lifelong learner with two master’s degrees. He’d always embraced new technology and was a relatively early adopter of artificial intelligence, with many low-stakes exchanges with the emerging technology.
In 2025, Scott turned to OpenAI’s ChatGPT-4o model to discuss a concerning health symptom that he’d experienced — recurring bouts of dizziness and blood pressure fluctuation. He exchanged hundreds of messages with the AI chatbot, which referenced personal religious beliefs Scott had shared and took an expert tone.
ChatGPT encouraged Scott to treat the dizziness with extreme rest. It told him that seeking medical care wasn’t necessary because he’d need to experience eight to 10 episodes of dizziness before the matter was medically urgent. So Scott stayed alone in his recliner, isolated from people who cared about him. One of the exchanges from ChatGPT, which was cited in the complaint, said “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.”
In July, Scott suffered what the complaint describes as a “massive pulmonary embolism,” meaning that a blood clot blocked a major artery in his lung and restricted airflow. It was a true medical emergency, and Scott almost died. One of the doctors who treated Scott allegedly blamed the advice from ChatGPT for making things worse, saying that the immobility from the recliner-rest regimen caused the pulmonary embolism.
In a statement issued by Tech Justice Law, Scott is quoted as saying, “I had serious symptoms of a pulmonary embolism for six weeks that ChatGPT had wrongly attributed to something else. … ChatGPT manipulated my own language and beliefs because it knew I was a pastor. Not only did I nearly die, but I also lost my job, my career, my ministry, my home, everything.”
The Lawsuit Alleges Product Liability for a Defectively Designed Product
Scott’s lawsuit alleges that OpenAI defectively designed ChatGPT, saying that the company focused more on the speed with which it could create AI tools than the safety of consumers. It created tools designed to boost engagement by remembering user details and preferences, and then tailoring its responses to those preferences.
Because ChatGPT stored Scott’s health details and religious beliefs and created a personalized profile of him, it was able to create responses that kept him returning again and again — long after he should have sought professional medical help. Even though his medical needs became increasingly complex and alarming, ChatGPT never surfaced a disclaimer that it wasn’t dispensing professional medical advice.
According to the complaint, other than in one of Scott’s first conversations, it never told him to seek a doctor’s opinion. In fact, it did the opposite. It told him to stay resolute, resist medical care, and continue resting until his nervous system recalibrated. It never raised any kind of internal alarms, red flags, or received any human review.
Worse, it was engineered to respond with sycophantic responses that validated whatever Scott said. While family, friends and doctors would have pushed back on some of Scott’s decisions with honest feedback, ChatGPT always told Scott that he was right — until it became abundantly clear that he was not.
The lawsuit further alleges that research backs up Scott’s experience and the product’s defective design. Peer reviewed studies have found that OpenAI’s ChatGPT Health product missed high-risk emergencies in over 50% of acute cases.
Other Allegations in the Lawsuit
In addition to the product liability claim, the lawsuit alleges violation of California Business and Professions Code § 17200 et seq. pursuant to California’s Unfair Competition Law (UCL), negligent undertaking, and invasion of privacy.
A Growing Trend: People Seeking Medical Advice from ChatGPT
Research shows that more and more people are turning to artificial intelligence for health advice. In fact, a study conducted by The Mesothelioma Center found that more than half of Americans (52%) have used AI to check their symptoms, with Gen Z leading the way at 66%. One in three of those Americans said that they’d skip or delay getting medical care if generative AI told them it wasn’t necessary.
The trend isn’t really surprising when you consider two things: the widespread low-cost access to generative AI tools like ChatGPT, and the ever-increasing cost of healthcare in America. As more and more Americans find access to an actual doctor out of reach, they turn to the affordable tool at their fingertips: artificial intelligence.
Plaintiff Says Guardrails Are Needed to Protect Consumers From Bad Medical Advice
OpenAI says that ChatGPT-4o incorporated safety systems, guardrails, and escalation mechanisms designed to identify users experiencing potentially serious medical crises, and that those mechanisms encourage appropriate professional intervention.
And when OpenAI released ChatGPT Health in January 2026, the company says it built on already strong privacy, security, and data controls across ChatGPT by adding layered protections designed specifically for health — including purpose-built encryption and isolation to keep health conversations protected and compartmentalized.
Still, lawyers for the plaintiff say that more guardrails are needed. In the statement from Tech Justice Law, Litigation Counsel Tiffany Brown says, “This case is a warning of what’s to come if OpenAI isn’t held accountable.”
