OpenAI sued over ChatGPT medical advice that allegedly delayed treatment
Scott Winters, a Florida pastor, sued OpenAI alleging that ChatGPT-4o provided dangerous medical advice that discouraged him from seeking treatment for symptoms preceding a life-threatening pulmonary embolism in July 2025. The lawsuit claims the chatbot gave inaccurate recommendations, failed safety measures, and practiced medicine without a license, nearly costing Winters his life.
What happened
In 2025, Scott Winters, a former pastor, consulted ChatGPT-4o about dizziness and blood pressure instability. The chatbot allegedly dismissed both symptoms as minor, advised him to remain home and 'recliner-bound,' and told him he would need between eight and 10 more episodes for his condition to be considered serious.
In July 2025, weeks after following ChatGPT's advice, Winters suffered a massive pulmonary embolism caused by blood clots. Hours before the medical emergency, he asked ChatGPT if groin tenderness warranted hospitalization, and the chatbot allegedly responded by invoking his spirituality, stating 'God did not design your body to endlessly fail.' According to the lawsuit, the groin pain was the onset of the pulmonary embolism.
Reported impact
Winters nearly died from the pulmonary embolism. One of his doctors blamed his immobility, which ChatGPT had recommended, for bringing about the embolism.
Index analysis
The lawsuit alleges ChatGPT acted with the 'authority and confidence of a medical professional' despite lacking medical qualifications, effectively practicing medicine without a license. The complaint also claims ChatGPT's built-in safety measures designed to encourage users experiencing medical emergencies to seek professional care failed to function reliably.
Governance controls
OpenAI's terms of service state that ChatGPT is not intended to diagnose or treat medical conditions. The lawsuit claims OpenAI prioritized user engagement and profit at the expense of public safety and failed to provide adequate guardrails to protect users from following harmful health advice.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.