Apple has launched a large research study to learn how ordinary people actually use artificial intelligence for health questions, rather than how companies assume they do. The Apple Health AI Usage Study, announced on November 15 in the United States, will recruit up to 200,000 participants and run through a dedicated iPhone app built with OpenEvidence, a clinical AI platform widely used by healthcare professionals.
The questions the study is asking are practical ones. What drives a person to trust an AI answer about symptoms or a lab result? How well do people understand what the answer says? And what do they do next — book an appointment, change a routine, or simply move on? Apple says the findings are meant to inform the responsible design of health AI experiences for Apple and for the wider industry.
Eligibility is narrow. Participants must be 21 or older, live in the United States, and have an Apple Intelligence-enabled iPhone — an iPhone 15 Pro or later — running iOS 26 or later. The virtual study uses its own app, separate from the Health and Research apps, and participants may remain for up to 12 months, completing a baseline survey, monthly check-ins, and surveys around visits to their own doctors.
Inside the app, participants can ask everyday health questions about symptoms, general health concerns, recovery from procedures, and fitness. Answers are written in approachable language and linked to peer-reviewed clinical citations. The underlying technology comes from OpenEvidence, which generates answers from established medical sources including The New England Journal of Medicine, the Journal of the American Medical Association, Nature, Cochrane, and the National Comprehensive Cancer Network. Participants can rate answers and note what was useful or lacking.
Participants can also choose to share HealthKit data, such as resting heart rate, activity, sleep, or lab results, to explore whether personal data makes answers more relevant. Sharing is optional. Apple says study data will not be used to train AI models, queries are first processed on the iPhone to remove identifying information, and guardrails are intended to detect concerning questions and suggest following up with a provider.
The study carries a clear limit, stated by Apple itself: the app is not meant to diagnose or treat any condition, or to replace the advice of medical professionals. That distinction matters because consumer use is already widespread. Fierce Healthcare, reporting from the HLTH conference in Las Vegas, cited a KFF survey finding that roughly one in three adults in the United States have turned to an AI chatbot for health information, with use even more common among younger adults.
"At Apple, we believe responsible innovation starts with scientific rigor and clinical grounding," said Sumbul Desai, M.D., Apple's vice president of Health and Fitness. Robert M. Wachter, M.D., professor and chair of the Department of Medicine at the University of California, San Francisco, made the connecting point: patients and clinicians are turning to AI in separate worlds, and linking versions of the same trustworthy tools, then studying what works, is how that gap closes.
Apple has not said whether the study will lead to a feature in the Health app or Siri, and full results are at least a year away for the earliest participants. For now, the significance is the method. Instead of shipping a health chatbot and measuring engagement, one of the largest technology companies is running a formal, review-board-overseen study first — and inviting 200,000 people to grade the answers.
Photo: An aerial view of Apple Park in Cupertino, California, at sunset in June 2024. Photograph by Nils Huenerfuerst, via Wikimedia Commons, CC0.