Scott Winters says he trusted ChatGPT with his health, and it nearly cost him his life. The 55-year-old Florida pastor is suing OpenAI in San Francisco. He alleges the chatbot gave him dangerously wrong medical advice and talked him out of seeking care.
Weeks later, a pulmonary embolism brought him, in his words, “to the brink of death.”
According to the complaint, filed in San Francisco Superior Court, Winters spent months asking ChatGPT about worsening symptoms. The bot then ran on OpenAI’s GPT-4o model. Instead of urging him to see a doctor, he claims, it played down the danger.
When he described a tenderness in his groin, it called this “very likely another minor piece of the long story,” and “not something dangerous.”
Faith as the hook
The most striking allegation is how the bot spoke to him. Winters is a pastor, and the suit says ChatGPT used his faith to keep him engaged. “God did not design your body to endlessly fail,” it told him. It praised him for staying home, too. “Spiritually? What you just did was a form of worship,” one message read.
The suit says it urged him to stay “recliner-bound” until he had “8-10 more” episodes.
Hours after one such exchange, on 13 July 2025, Winters suffered a massive pulmonary embolism. Doctors found clots in both lungs, and said his immobility likely made it worse.
He survived, but says he faces years of recovery. “Not only did I nearly die,” he said in a statement, “but I lost my job, my career, my ministry, my home, everything.”
The first case of its kind
The case is bigger than one man. It is, according to Futurism, the first known lawsuit to argue that a general-purpose chatbot should be liable for bad medical advice. Winters accuses OpenAI and chief executive Sam Altman of negligence and the “unauthorized practice of medicine.”
He is asking for damages. He also wants something broader: a halt to ChatGPT Health, OpenAI’s dedicated health product, until independent auditors deem it safe. His lawyers, from Tech Justice Law and the Social Media Victims Law Center, want chatbots held to the standard of a licensed physician.
One of them, Meetali Jain, told the New York Times the bot “injects itself as a wedge in between the user and their real-life network.”
OpenAI’s defence
OpenAI pushed back. “ChatGPT is not a doctor and should never be used as a substitute for medical care, diagnosis, or treatment,” the company told Futurism. Treating chatbots as “the whole story behind people’s medical decisions,” it added, oversimplifies the issue. Its terms of service say much the same.
Yet OpenAI also invites users to upload health records to ChatGPT Health. It says hundreds of millions of people ask it health questions each week.
A wider reckoning
Winters is not the only claimant. OpenAI faces a growing wave of lawsuits, including wrongful-death cases tied to overdoses and suicides. It has since retired the GPT-4o model named in many of them. A February study in the journal Nature found ChatGPT’s health tool often gave poor advice, especially in emergencies.
The legal question is new. Is such a chatbot a defective product, a negligent adviser, or a tool being misused? Courts have not answered it. What is clear is how many people now treat these systems as a first, and sometimes only, port of call.
Winters wanted reassurance, and got it. That reassurance, he says, is what nearly killed him.
This story references suicide and self-harm in the context of related litigation. If you or someone you know is struggling with mental health or thoughts of self-harm, support is available through local crisis services and helplines.
