OpenAI has made Health in ChatGPT available to all eligible users in the United States, the company said on July 23, 2026, turning a pilot that began in January into a permanent feature for logged-in adults.
Once a user connects their Apple Health data or medical records — including from Epic, Oracle Health, and One Medical — ChatGPT Health can draw on lab results, medications, sleep patterns, and activity history to help track trends over time, prepare questions before a doctor’s visit, or explain how daily habits relate to overall health. The feature is open to users 18 and older on the Free, Go, Plus, and Pro tiers, on web and iOS.
Built into every conversation, not just one tab
During the pilot, health context only applied inside a dedicated Health section. OpenAI said it found that roughly 70% of health-related questions came up outside that section — in chats about travel plans, meal ideas, or workouts — so the wider release lets the assistant apply a user’s health context across any conversation, once permission is granted.
Medical-record connectivity runs through Baltimore-based b.well Connected Health, a partner OpenAI says links to more than 2.2 million US providers and 320 health plans and labs. The feature runs on OpenAI’s GPT-5.6 model family, with GPT-5.6 Sol handling more complex health questions; the company says it worked with practicing physicians to benchmark the models before expanding access.
Data won’t train the models, OpenAI says
OpenAI says connected health data, and any conversations that draw on it, are encrypted both in transit and at rest and are never used to train its foundation models or serve ads. More than 300 million people already ask ChatGPT a health-related question in a typical week, the company said, up from roughly 230 million when the pilot launched in January. OpenAI continues to describe the assistant as a research aid rather than a clinician, saying it “is not a substitute for professional medical care.”
The launch adds a specific, permission-gated case to the broader question of what happens to personal data once it enters an AI tool — one that has followed AI’s growing footprint in healthcare and medicine more broadly.