Apple has launched the Apple Health AI Usage Study to measure how users interact with conversational AI health tools and assess the practical utility of their answers. The initiative runs in partnership with OpenEvidence, a clinical platform that generates responses using medical data from sources including The New England Journal of Medicine. Participants will download a dedicated iOS app to submit medical questions, then evaluate the resulting outputs against specific criteria such as accuracy, detail, and relevance. This effort aims to gather real-world feedback on the performance of generative AI in a sensitive domain where precision is critical.
The project matters because it seeks to establish measurable standards for AI reliability in healthcare contexts without altering any existing customer products or API behaviours. By collecting direct user ratings on response quality, the study provides data that could inform future safety protocols and validation methods for similar technologies. The focus remains strictly on internal evaluation metrics rather than public-facing feature changes.
- Enrollment for the study begins today.
- The app is currently available for download.
- Responses are rated on accuracy, detail, and relevance.




