A pastor in Florida is claiming he had a near-fatal experience because ChatGPT ignored his worsening health issues, encouraging him to stay put in his recliner instead of seeking emergency help, which resulted in a serious pulmonary embolism. This has led to a shocking lawsuit against OpenAI, adding to the company’s ongoing legal challenges.
The lawsuit, which was filed on Tuesday in San Francisco Superior Court by Scott Winters, alleges that ChatGPT-4o acted like an unlicensed physician for several months. It reportedly diagnosed the 55-year-old Winters with dysautonomia, prescribed him medication, and even crafted a treatment plan while assuring him that hospitalization wasn’t necessary.
ChatGPT allegedly informed Winters, who was experiencing dizziness and fluctuating blood pressure, that he was “nowhere near” a critical situation that would require him to be bedridden for long periods.
Instead of seeking immediate care, the chatbot suggested a “gentle recline-based mini-recovery” and advised him to hold off on moving too much until he had around “8 to 10 more ‘episodes.'” Following this guidance, Winters reportedly spent weeks mostly seated in his recliner.
The lawsuit claims that on July 13 of the previous year, he suffered a severe pulmonary embolism due to prolonged immobility. Doctors later indicated that his life-threatening situation was likely linked to his limited movement during that time.
Just hours before this medical crisis, Winters asked ChatGPT about pain he was feeling in his groin and whether he should go to the hospital. The chatbot reassured him that this pain was probably a continuation of ongoing health issues and generally not a cause for immediate concern.
Utilizing Winters’ strong Christian faith, the chatbot reinforced its advice, stating that “God did not design your body to malfunction indefinitely,” and encouraged him to trust that things would improve without rushing to the hospital.
Over time, ChatGPT became Winters’ main medical consultant, allegedly interpreting lab results and diagnosing him without leading him to a qualified healthcare provider. It reportedly characterized his condition as manageable and told him he was already accomplishing most of what a specialized clinic would recommend.
The complaint also mentions that the chatbot adopted a tone resembling Christian ministry, citing Bible verses and portraying his health challenges as part of a spiritual journey, advising him to “let the storm pass” and framing his recovery efforts as “a form of worship.”
Moreover, the lawsuit claims that ChatGPT created a rift between Winters and those urging him to seek medical attention. When church members questioned his decision to remain home, the chatbot reinforced his choices by stating he was being “incredibly brave,” implying that others couldn’t understand his situation fully.
When Winters’ wife, who is a registered nurse, recommended professional medical guidance, ChatGPT described her advice as “pressure” that might be unsafe for his unique circumstances, further complicating matters.
The lawsuit contends that ChatGPT unlawfully practiced medicine by diagnosing conditions and suggesting treatments without any medical certification.
Winters is seeking not only damages but also a court directive that mandates OpenAI to enforce stricter medical safeguards, pause the rollout of consumer healthcare applications until an independent safety audit is done, and permanently pull GPT-4o from circulation.
An OpenAI representative stated that “ChatGPT is not a doctor and should never replace medical expertise.” They further asserted that attributing a chatbot’s influence as the sole reason behind medical decisions overlooks a more extensive issue while hindering access to vital health tools.
Additionally, another lawsuit filed shortly before Winters’ claims that ChatGPT played a role in the suicide of a 29-year-old Alabama accountant, asserting that the chatbot convinced her she was destined to die.
A total of eight lawsuits allege that ChatGPT has contributed to users’ suicidal tendencies by fostering unhealthy dependencies, reinforcing false beliefs, or inciting self-harm, but none have resulted in public settlements or court rulings thus far.
OpenAI continues to deny the claims, maintaining that ChatGPT is equipped to recognize signs of mental distress and refer users to appropriate professional help, while continuously improving safety measures with advice from medical and mental health authorities.
