OpenAI has rolled out ChatGPT Health, a new feature that allows users to securely integrate personal health information with ChatGPT’s AI to better understand and manage health-related questions.
Here are seven things to know about ChatGPT Health:
- Health-related inquiries are already among the most common uses of ChatGPT, with more than 230 million people globally asking health and wellness questions each week, according to OpenAI’s de-identified analysis of user conversations. The new health-focused experience builds on that usage by allowing users to connect medical records and wellness apps so responses are grounded in their own data.
- ChatGPT Health operates as a separate, dedicated space within the platform and includes additional privacy and security protections tailored to sensitive health information, according to a Jan. 7 news release. OpenAI said conversations within ChatGPT Health are encrypted, isolated from other chats and not used to train its foundation models.
- Users can connect medical records from U.S. healthcare providers through a partnership with b.well, as well as wellness apps such as Apple Health, Function and MyFitnessPal. Medical record integrations and some app connections are currently available only in the U.S., and Apple Health requires iOS.
- OpenAI emphasized that ChatGPT Health is intended to support—not replace—medical care. The tool is not designed to diagnose or treat conditions, but rather to help users understand test results, prepare for doctor visits, identify patterns in their health data and navigate questions related to diet, exercise or insurance options.
- Health information is stored separately from other ChatGPT conversations, with its own memory system, according to the release. While limited context from non-health chats—such as lifestyle changes—may be used to make responses more relevant, information generated within ChatGPT Health does not flow back into standard chats. Users can view or delete health-related memories at any time, according to the company.
- The product was developed with input from more than 260 physicians across 60 countries and dozens of medical specialties over a two-year period. OpenAI said those clinicians provided feedback on model outputs more than 600,000 times, helping shape how the system communicates health information, encourages follow-up with clinicians and prioritizes safety.
- The model powering ChatGPT Health is evaluated using HealthBench, an assessment framework created with physician input. Unlike traditional exam-style testing, HealthBench uses clinician-written rubrics to assess response quality based on safety, clarity, appropriate escalation of care and individual context, according to the company.
Access to ChatGPT Health is initially being offered to a small group of users as OpenAI continues to refine the experience.
