ChatGPT Health is launching as a dedicated health-focused version of ChatGPT that combines personalized health data with stronger privacy and security controls to support not replace conversations with clinicians.
The new experience isolates health chats, encrypts data with additional protections, and allows users to securely connect medical records and wellness apps so responses are more tailored to their real health history.
What ChatGPT Health offers
ChatGPT Health is designed to help people understand their health information, prepare for appointments, interpret test results, and manage everyday wellness decisions.
Users can securely connect data from sources like Apple Health, Function, MyFitnessPal, and U.S. medical records (via partner b.well) so the assistant can ground answers in their lab results, activity trends, medications, or insurance details.
The experience focuses on explanation, context, and planning such as clarifying lab reports, helping organize questions for a doctor, or outlining lifestyle strategies rather than diagnosing or prescribing treatment.
Health is positioned as a companion that supports informed conversations with healthcare providers, not a replacement for professional care.
Health operates as a separate space within ChatGPT where conversations, connected apps, files, and memories are stored and processed separately from other chats.
Health-specific information does not flow back into general ChatGPT conversations, and non-Health chats cannot access Health files or memories, helping compartmentalize sensitive medical data.
Enhanced privacy and security protections
All ChatGPT conversations are encrypted at rest and in transit, but Health adds purpose-built encryption and isolation specifically for health data, reflecting its higher sensitivity.
Crucially, conversations in Health are not used to train foundation models, and users can view or delete Health memories and disconnect apps or record access at any time through settings.
When users choose to link health records or wellness apps, the system is designed so only the minimum necessary data is accessed and only with explicit permission.
Apps approved for use in Health must meet additional privacy and security requirements, undergo specific reviews, and clearly disclose what data they can access when first connected.
Users can revoke app access at any time, after which the third party immediately loses the ability to retrieve new data through Health.
Multifactor authentication can also be enabled on the account to further reduce the risk of unauthorized access to sensitive health chats and connected data.
ChatGPT Health was developed over two years in close collaboration with more than 260 physicians across 60 countries and many specialties, who collectively provided feedback on hundreds of thousands of model responses.
Their input shaped how the system prioritizes safety, when it should encourage urgent follow-up with clinicians, and how to balance clear language with medically accurate nuance.
OpenAI created HealthBench, a clinical evaluation framework with physician-written rubrics that measure safety, clarity, escalation, and contextual appropriateness instead of just exam-style correctness.
This framework helps ensure ChatGPT Health performs well on real-world tasks like explaining lab values, summarizing care instructions, or interpreting wearables data, while reinforcing that its role is to support not replace healthcare providers.
ChatGPT Health is rolling out first to a limited group of early users who join the waitlist, with eligibility for people on Free, Go, Plus, and Pro plans outside the EEA, Switzerland, and the UK.
As the experience is refined, access will expand to all users on web and iOS, with some integrations such as U.S. medical records and certain apps initially limited to the United States and requiring iOS for Apple Health connections.
Follow us on Google News, LinkedIn, and X to Get Instant Updates and Set GBH as a Preferred Source in Google.