ChatGPT Health Connects Medical Records and Apple Health
OpenAI’s new health experience can ground answers in personal records and wearable data, making its privacy controls as important as its medical reasoning.
OpenAI began rolling out ChatGPT Health on July 23, giving eligible U.S. users a way to connect medical records and Apple Health data to conversations. The most important change is not that ChatGPT can discuss a lab result. It is that the answer can now draw on the person’s own medications, previous results, visits, sleep and activity—with permission.
The rollout starts with logged-in users aged 18 and older across Free, Go, Plus and Pro plans. It is available on the web and iOS, but not yet in Codex. OpenAI says users can connect Apple Health, supported records from U.S. hospital systems, One Medical and Function Health through the new Health area.
What changes for users
Personal context can make a health conversation less generic. A user could ask how a new result compares with earlier tests, review changes since an appointment or prepare questions for a clinician without gathering the same background every time. Health also acts as a dashboard for synced information and past health conversations.
That convenience comes with a practical responsibility: connected information may be incomplete or out of date. OpenAI advises users to review conditions and medications and to check important details against the original source. A medication that has been stopped, for example, may still appear in a connected record.
The privacy bargain
OpenAI says connected medical records, Apple Health information and conversations that use that information are not used to train its foundation models or target advertising. The company also says health information receives additional encryption protections.
Access is permission-based. By default, ChatGPT asks before using connected records or Apple Health data in a response. Users can approve a single request or allow continuing access, and can change that choice later.
Disconnecting a source starts deletion of its synced data from OpenAI’s systems, which the company says occurs within 30 days. There is an important distinction: information already placed in ordinary ChatGPT conversation history remains there until those conversations are deleted.
What it cannot promise
OpenAI says ChatGPT can still make mistakes and does not replace the judgment of a qualified medical professional. That caveat becomes more consequential, not less, when an answer feels personal and informed. Connected context can help people ask better questions; it does not turn probabilistic software into a clinician.
The product test is therefore bigger than model performance. ChatGPT Health asks users to exchange highly sensitive context for more useful answers. Whether that bargain earns lasting trust will depend on clear permissions, predictable deletion and restraint when the model is uncertain.
Status
Confirmed. Availability, supported connections and privacy controls are documented by OpenAI. Performance and safeguard descriptions remain first-party claims rather than an independent medical evaluation.
Sources
Update note: Last reviewed 2026-07-24. We will revise this post if primary channels update.
Sources
- OpenAI — Launching Health in ChatGPT — official
Drafted with AI assistance from source briefs; reviewed for citation completeness and label accuracy.