OpenAI Unveils ChatGPT Health to Integrate Medical Records and Wellness Data

OpenAI has announced the launch of ChatGPT Health, a dedicated environment within its flagship artificial intelligence platform designed to help users manage personal health information. The new interface allows for the direct integration of medical records and data from wellness applications, marking a significant step toward personalized AI health assistance.

Personalized Insights Through Data Integration

According to OpenAI, health-related queries are among the most frequent use cases for its platform, with an estimated 230 million people globally asking health and wellness questions every week. ChatGPT Health aims to move beyond general advice by grounding conversations in a user’s specific medical history and physiological data.

U.S. users can now connect the platform to over 2.2 million healthcare providers through a partnership with b.well, a health data network. This integration enables the AI to analyze lab results, clinical summaries, and vaccination records. Additionally, the service supports data syncing from Apple Health, MyFitnessPal, and Function, allowing the AI to factor in sleep patterns, nutrition, and cardiovascular metrics when responding to user prompts.

Siloed Privacy and Data Security

Recognizing the sensitivity of medical data, OpenAI has implemented a compartmentalized architecture for the new service. Health conversations and synced files are stored in a dedicated space, separate from standard user interactions. While the AI may use general context from outside the Health tab to improve relevance, data generated within the Health environment is prohibited from flowing back into non-health chats.

Critically, the company stated that conversations and records within ChatGPT Health will not be used to train its foundational models. The service also employs purpose-built encryption and isolation protocols. To further secure accounts, the company is encouraging users to enable multi-factor authentication.

Clinical Oversight and Limitations

The development of ChatGPT Health involved a two-year collaboration with more than 260 physicians across 60 countries. This medical network provided feedback on over 600,000 model outputs to refine how the AI communicates clinical information and when it should escalate concerns to a human professional. The model is evaluated against HealthBench, a proprietary assessment framework built to prioritize safety and clarity over generic accuracy.

Despite these clinical safeguards, OpenAI emphasized that the tool is intended for support rather than diagnosis or treatment. The interface is designed to help users interpret complex medical jargon, prepare for doctor appointments, and identify long-term patterns in their wellness data, rather than serving as a replacement for professional medical care.

Availability and Regional Restrictions

The service is initially rolling out to a small group of early users on ChatGPT Free, Plus, Pro, and Team plans. While OpenAI plans to expand access to web and iOS platforms in the coming weeks, the feature is currently unavailable to users in the United Kingdom, Switzerland, and the European Economic Area (EEA), likely due to regional data protection regulations. Medical record integrations remain exclusive to the United States at this stage.