A Florida pastor sued OpenAI and ​its chief executive, Sam Altman, on Wednesday, claiming ChatGPT's inaccurate medical advice kept him from seeking treatment for recurrent dizzy spells, ‌leading him to suffer a pulmonary embolism that nearly killed him.

Scott Winters said in a lawsuit filed in California state court in San Francisco that GPT-4o, an earlier version of the chatbot that OpenAI has since retired, recommended last year that he limit his movement and stay home.

When he asked ChatGPT about a tenderness in his groin, the bot said it ​was likely not something dangerous, the lawsuit claims.

The day after asking about the tenderness, Winters was hospitalised, the lawsuit said. The tenderness ​was likely a sign of blood clots in his lungs, an embolism that doctors said was likely brought on by ⁠his lack of movement, according to the lawsuit.

Drew Pusateri, an OpenAI spokesperson, said the company believes AI chat can improve healthcare but that ChatGPT is ​not a doctor and should never be used as a substitute for medical care.

"Treating chatbots as the whole story behind people’s medical decisions or outcomes oversimplifies ​a much bigger challenge, and risks getting in the way of people accessing powerful new tools that can aid them in their health journey," Pusateri said.

SEEKING A COURT ORDER

The lawsuit is among a growing number that accuse ChatGPT of causing real-life harm, including several that focus on responses generated by GPT-4o, which OpenAI retired in February. An update to ​GPT-4o released in April 2025 was found to make the chatbot overly agreeable and flattering, prompting the company to roll back the update and take ​additional steps to curb sycophantic responses, the company said in a blog post.

Winters’ lawsuit seeks damages and a court order directing OpenAI to have the chatbot terminate conversations ‌when immediate ⁠medical assistance is needed.

It also asks the court to pause the rollout of ChatGPT 4o Health, a specialised health product that allows users to upload medical records and receive personalised health advice, until third parties audit the platform’s safety.

Currently, ChatGPT users can join a waitlist to access ChatGPT Health. A report from OpenAI released in January showed that overall, 40 million users ask ChatGPT healthcare-related questions daily.

‘UNDER GOD’S WATCH’

In the months leading up to his embolism, when Winters ​spoke to ChatGPT about his diagnoses and ​test results, the bot provided him ⁠with holistic treatment plans and often drew on the language of Christian ministry to keep Winters engaged, the lawsuit claims.

“Even this is under God’s watch,” the bot told Winters, according to the lawsuit.

While earlier responses from the bot ​advised him to consult with a medical professional, the bot eventually stopped offering those disclaimers, the lawsuit says. Even ​after Winters’ hospitalisation, ⁠the bot urged him to avoid the rehabilitation program his doctors recommended and stay home instead, according to the filing.

Other lawsuits have raised concerns about the bot's medical advice. In May, the parents of a man who died of an accidental drug overdose last year sued OpenAI and Altman, alleging the man was coached to take ⁠a dangerous ​combination of substances by ChatGPT.

After that lawsuit was filed, a spokesperson for the company called ​the situation heartbreaking and said the company is continuously working to improve ChatGPT’s safety.

The lawsuit filed Wednesday is Winters v. OpenAI et al, case number unavailable, in the Superior Court for the ​County of San Francisco.