Introducing Apple and OpenEvidence’s Medical AI Study
Apple, in collaboration with OpenEvidence, is initiating a 12-month research project focused on exploring how artificial intelligence can provide medical advice through a dedicated app. This study will involve up to 200,000 iPhone users across the United States, allowing participants to ask health-related questions and receive responses grounded in scientific and clinical evidence.
How the App Works
The specially designed application pulls information from reputable medical journals, including:
- The New England Journal of Medicine
- Journal of the American Medical Association
- Nature
By translating complex medical information into easy-to-understand language, the app aims to be accessible to a broad audience. Users will also be able to rate answers, provide feedback, and participate in monthly surveys to help improve the experience.
To qualify for the study, participants must be US residents aged 21 or older, own an iPhone 15 or newer, and run iOS 26 or later. They will need to activate Apple Intelligence features and may choose to share HealthKit data such as resting heart rate, physical activity, sleep patterns, and lab results. This data will support personalized health guidance, preparation for doctor visits, and recommendations on when to consult medical specialists for concerning symptoms.
Importantly, all shared information is anonymized before being securely stored on a third-party cloud service. The data will not be used to train AI models, and OpenEvidence commits to deleting all query and response information within 24 hours. The app is currently available for download in the US. Apple has not announced plans to integrate this functionality with the Health app or Siri voice assistant.
Sumbul Desai, a spokesperson for the project, commented: "This study aims to understand how people perceive AI-generated medical responses, how well they comprehend them, and how they evaluate their usefulness in everyday life."
This initiative represents a major advancement in applying artificial intelligence to healthcare, potentially increasing access to reliable medical advice for many. The findings will also provide valuable insights to enhance patient-provider communication and improve the accuracy and trustworthiness of AI-driven health recommendations.