Ready to let AI pore over your medical records? Claude and ChatGPT are already doing it, and now Microsoft’s Copilot is ready to review your chart.
Copilot Health is a “separate, secure space” within the Copilot app that “makes sense of your information” from a variety of health-related sources, from your Apple Watch and Oura ring to uploaded lab results, and medical records from your doctor, Microsoft says.
As with ChatGPT Health and Claude’s health integrations, Copilot Health isn’t designed to “replace” your doctor, Microsoft carefully notes. Instead, it’s intended as a way to sort through reams of medical data and turn it “into a coherent story,” allowing you to “arrive prepared, with the right questions” for your actual doctor visits.
For now, Copilot Health is available only in the U.S. for users 18 and over, and you’ll have to join a waitlist before trying it out.
Once you’re signed up with Copilot Health, you’ll be able to pipe in your medical records via HealthEx, a healthcare integration that connects with more than 50,000 U.S. hospitals and offers details about your hospital visits, medications, and test results.
Copilot Health can also gather health intel from Apple Health, Oura, and Fitbit, good for giving the AI an overview of your daily activity, sleep patterns, and vitals.
Your Copilot Health data and chats will be encrypted and “isolated” from the main Copilot app, Microsoft promises, and you’ll be able to delete your medical info at will or revoke access to your health integrations at any time.
Copilot Health is jumping into a small but rapidly growing market of AI chatbots that can scan your medical records and offer natural-language summaries and health advice.
ChatGPT Health, which launched back in January, can also sift through your Apple Health data, and it also works with MyFitnessPal and uploaded medical records.
Then there’s Claude for Healthcare, which works with Apple Health, Android Health Connect, and the same HealthEx integration as Copilot Health.
All the big AI providers are positioning their health services as methods for making sense of your medical records rather than ways to replace a human doctor.
But while Copilot Health and other AI chatbots could be helpful for translating confusing medical results or the jargon from a lab tech’s X-ray notes, the accuracy of their overall health advice has come under increasing scrutiny.
Take a recent study published by Nature Medicine, which found that ChatGPT Health “under-triaged” more than half of simulated emergency scenarios fed to it by medical researchers, meaning it frequently failed to recognize health situations in which the user should stop chatting and call 9-1-1.
On the other hand, some doctors argue that health-oriented AI chatbots can offer “useful information” when “used responsibly,” and that the alternative “often is nothing, or the patient winging it.”