Microsoft’s Copilot Health can connect to your medical records and wearables
The new chatbot can analyze your lab results and find doctors who take your insurance.
Microsoft has launched Copilot Health, a dedicated, secure environment within its Copilot AI assistant designed for personal health inquiries. The chatbot can connect to a user's medical records from over 50,000 US hospitals and healthcare organizations through HealthEx and import lab results via Function. It also integrates with data from more than 50 wearable devices, including Apple Watch, Oura ring, and Fitbit, to display metrics like step count alongside upcoming appointment reminders. Microsoft emphasizes this is not for diagnosis but for helping users understand their health data and navigate the healthcare system, such as finding doctors based on specialty, location, and accepted insurance plans.
Responses in Copilot Health are designed to be reliable, with citations to sources and expert-written answer cards from Harvard Health. Microsoft states chats are isolated from general Copilot use and kept under additional privacy controls, with health data not used to train AI models. Users can delete data or disconnect sources at any time. Notably, Microsoft currently does not offer a HIPAA-compliant version for this consumer-facing product, unlike competitors OpenAI's ChatGPT Health and Anthropic's Claude for Healthcare. Dr. Dominic King, VP of health at Microsoft AI, stated HIPAA isn't legally required for this direct-to-consumer experience but indicated the company would announce updates on its 'HIPAA controls.' The feature is rolling out in phases, with users able to join a waitlist for access.
- Connects to medical records from 50,000+ US hospitals via HealthEx and 50+ wearables like Apple Watch and Fitbit.
- Helps users decipher lab results and find in-network doctors using real-time US provider directories.
- Chats are isolated, not used for AI training, but the consumer version is not currently HIPAA-compliant.
Why It Matters
It puts AI-powered health navigation and data synthesis directly in consumers' hands, though privacy standards differ from clinical tools.