OpenAI has started rolling out Health in ChatGPT to logged-in users aged 18 and older in the United States on web and iOS, across the Free, Go, Plus and Pro plans. With explicit user permission, ChatGPT can securely connect to Apple Health and supported medical records so it can provide context-aware, personalized responses based on the information you choose to share.
Why this feature was introduced
More than 300 million people consult ChatGPT each week with health-related questions, from interpreting lab results and preparing for appointments to understanding clinicians’ notes and building healthier routines. That information is often scattered across patient portals, apps and wearables, making it hard to form a complete picture. Health aims to gather that context and make it easier to act on.
What Health does in ChatGPT
- With permission, ChatGPT can compare a new test result to prior tests, summarize what has changed since a previous visit, or explore links between sleep, activity and workouts and a user’s routine.
- The feature is intended to support users’ understanding and decision-making, not to replace medical professionals.
Model improvements and performance
OpenAI says it is improving underlying models to reason across complex health details, explain uncertainty, ask for missing context and recognize when escalation to professional care may be necessary. GPT‑5.5 Instant extended advanced health capabilities to free users, while GPT‑5.6 Sol — available to paid users — is described as delivering stronger performance on tasks that require reasoning across complex details. According to OpenAI, every GPT‑5.6 model outperformed GPT‑5.5 on the HealthBench Professional evaluation.
Clinical testing
Hundreds of physicians participated in testing Health, helping develop realistic scenarios and rubrics to evaluate accuracy, safety, communication, context-awareness, completeness and appropriate escalation. Despite progress, OpenAI cautions that ChatGPT can still make mistakes and should not replace qualified medical judgment; users are encouraged to verify important information and consult their healthcare providers.
Privacy and security safeguards
- Health is built with layered privacy and security protections. All ChatGPT conversations are encrypted at rest and in transit, and information connected in Health receives additional encryption.
- Connected medical records and Apple Health data, and conversations that use them, are not used to train OpenAI’s foundation models and are not used for ad targeting, regardless of a user’s model-training setting. Other conversations follow the user’s standard ChatGPT model-training settings, which can be changed in Settings > Data Controls.
- By default, ChatGPT asks for permission before using connected Health information. Users can approve each request or choose to always allow access; this setting can be changed at any time in Settings > Plugins > Health.
- Users can disconnect accounts at any time in Health > Accounts; when they do, data synced from that source is deleted from OpenAI’s systems within 30 days. Information already included in conversation history remains until the user deletes those conversations.
Memory and sensitive sharing
Memories remain under the user’s control: memories can be created from Health conversations, but they are not created automatically from connected medical records or Apple Health data. Users can avoid creating long-term memories by using Temporary Chat or turning memory off in Settings > Memory. For actions that might disclose Health information via connected plugins (for example, sending a training plan to a running partner based on Apple Health data), additional safeguards and confirmations are applied.
How to start and manage connected data
Open Health from the left sidebar or the More menu in ChatGPT and select “Get started” to begin securely connecting your health information. After connecting, syncing may take a few minutes; users can add more accounts or return later. If you use apps or devices that share data with Apple Health, make sure sharing is enabled — ChatGPT can use the information those apps expose through Apple Health when you allow it. Available data vary by app and metric; some proprietary scores may not transfer.
Users can review conditions and medications after connection, remove outdated items and add missing details such as family history. Because connected information may not always be complete or current, users should tell ChatGPT when something changes and verify important details against original sources.
Using Health in regular conversations
Based on early testing, OpenAI found that more than 70% of health-related conversations occurred outside a dedicated health space. For that reason, users can allow ChatGPT to draw on connected Health information across general conversations, without starting inside a separate Health workspace. When permitted, ChatGPT can consider medications, lab results, recent visits, sleep and activity alongside goals and context the user provides — for example, accounting for a dietary restriction when suggesting a restaurant or considering a recent injury when planning lower-impact activities. Users can also explicitly include health context by adding @Health to a message.
Limitations and final notes
OpenAI reiterates that the system can still make errors and does not replace clinical care. Connected data may be incomplete or out of date; users should update or correct details and consult healthcare professionals for important medical decisions.
In short, Health in ChatGPT brings the option to securely integrate Apple Health and supported medical records for U.S. users, with layered privacy protections, user-controlled permissions, clinical testing and improved models aimed at making AI-assisted health conversations more context-aware and useful.
