In any chat the tool can now draw on linked records, and it arrives a day after a Florida pastor sued OpenAI, claiming the chatbot told him not to see a doctor.
OpenAI has started opening its ChatGPT health feature to signed-in adults in the United States, letting them connect Apple Health and supported medical records so the chatbot can answer questions drawn from their medications, appointments, lab results, sleep and activity. OpenAI, through its official account, said on July 23 that it would securely link those records to read them in context and follow what has changed.
On iPhone, people can link Apple Health along with services such as Function and MyFitnessPal. They can also pull in medical records held by hospital providers like Epic and Oracle Health, plus services such as One Medical and Function Health.
The bigger shift is where that data can be used. Until now, health-related answers were confined to a dedicated hub OpenAI piloted in January 2026; with that limit lifted, ChatGPT can draw on a person's linked records in any conversation, whatever the subject. The company says its testing found that 70% of health activity was already happening outside that hub.
OpenAI says weekly health-related questions to ChatGPT have climbed to about 300 million, up from roughly 230 million in January.
On the privacy question that connecting medical data raises, OpenAI pledges that the linked data and the conversations relying on it will not be used to train its foundation models or to target ads. It points to safeguards including added encryption and permission prompts before each use.
The rollout lands a day after a pastor in Florida sued OpenAI, claiming ChatGPT advised him against seeing a physician — guidance he says nearly killed him. In reply to the suit, OpenAI cited its terms of service, which state that the product is not meant for diagnosing or treating any health condition, and it told The New York Times it is working to make its health- and medicine-related answers safer.
Alongside the launch, OpenAI urges people to fact-check what ChatGPT tells them and to base medical decisions on professional advice, and its terms warn against relying on the chatbot to diagnose or treat any health condition.