OpenAI has launched ChatGPT Health, a product that allows users to connect their medical records and wellness apps to the AI chatbot and get personalized health guidance.
The feature, unveiled on Wednesday, creates a separate space within ChatGPT for health questions and discussions, where users can collect data from their connected health apps, such as fitness apps, as well as store their health files.
Users can also connect to their electronic medical records through a partnership with B.well Connected Health, OpenAI says. ChatGPT, then, does not have a direct integration with the MyChart patient records app from Epic, for example, but lets individual users make requests for their patient record data through integrations built by B.well. In practice, the patient will see a login through their provider’s portal, which often is a MyChart login page to authenticate their account.
In addition, users can connect ChatGPT Health to wellness apps including Apple Health, Peloton, MyFitnessPal, and Function Health.
OpenAI say more than 230 million people globally already ask health and wellness questions on ChatGPT each week. The company says it developed ChatGPT Health over two years in collaboration with more than 260 physicians who have practiced in 60 countries, collecting over 600,000 pieces of feedback on model outputs.
The new product gets its intelligence from a specialized OpenAI health model. In collaboration with doctors, the company also created an evaluation tool called HealthBench, which it uses to test the health model.
Data in ChatGPT Health is protected using purpose-built encryption, OpenAI says, and health conversations in the Health space are not used to train OpenAI’s models. But privacy advocates remain concerned about the risks of sharing personal health data within a chatbot setting.
“While OpenAI says that it won’t use information shared with ChatGPT Health in other chats, AI companies are leaning hard into personalization as a value proposition,” says Andrew Crawford, a senior counsel at the Washington, D.C.-based Center for Democracy and Technology, in a statement. “Especially as OpenAI moves to explore advertising as a business model, it’s crucial that separation between this sort of health data and memories that ChatGPT captures from other conversations is airtight.”
OpenAI says health consumers can use ChatGPT Health to prepare for doctor appointments, understand clinical test results, get diet and exercise advice, and evaluate insurance options based on their healthcare patterns. ChatGPT Health is not approved by the Food and Drug Administration, so it’s not to be confused with real clinician diagnosis and treatment.
Right now, the new feature is available only to a small group of ChatGPT subscribers and free users. OpenAI plans to expand access and make its Health space available to all users on the web and iOS in the coming weeks. (You can sign up for the waitlist to request access.)
Fidji Simo, OpenAI’s CEO of applications, wrote in a blog post that she personally used ChatGPT to flag a potential medication conflict during a hospital stay last year, calling the experience an example of AI’s potential in healthcare.
“Because I’ve been dealing with a chronic illness for years, I had already uploaded a lot of my health records into ChatGPT,” Simo writes. “I asked whether I should be taking this antibiotic, given my medical history, and ChatGPT flagged that this particular antibiotic could reactivate a very serious infection I’d had a couple of years prior.”
In the big picture, ChatGPT Health represents yet another front in OpenAI’s growing platform war with legacy tech players such as Apple, Google, and Meta.