OpenAI has just announced the launch of ChatGPT Health, a specialized high privacy environment within ChatGPT designed to help users manage their personal health data. Health prompts have already been used by millions and ChatGPT Health aims to centralize scattered medical information, preparing them for doctor appointments and provide an overall view of their health.

The app acts as a compartmentalized space. It can pull conversations from your regular chats but information from the health space is never directed back towards your standard conversations. Conversations within Health are never used to train OpenAI’s foundational models.

Integrate with real world data. Users can securely connect their medical records and health apps to provide the AI with personalized context. Through a patnership with b.well, the largest and most secure network of live, connected health data for U.S. consumers, US users can connect their records from medical providers. Integrations include Apple Health, My Fitness Pal and Function to help interpret data from wearables and nutrition logs. Health can summarize lab results, compare insurance options based on your history and suggest questions to ask your doctor.

The tool was built over 2 years, in collaboration with 260 physicians across 60 countries. These professionals contributed over 600,000 pieces of feedback to the AI. Ensuring the the app prioritizes safety and clinical accuracy, communicates clearly without oversimplifying and knows when to urgently suggest a follow up with a human physician.

OpenAI emphasizes that the app is more of a support tool rather than a replacement for actual diagnosis. It is designed to help users assess their health over time rather than just during acute illnesses.

The feature is rolling out first to a small group of ChatGPT Free, Plus, Team, and Enterprise users outside of the UK, EEA, and Switzerland. It will eventually expand to all web and iOS users. Medical record integration remains limited to the U.S. for now.