Apple wants 200,000 iPhone users to test a health AI built on OpenEvidence
Apple is recruiting up to 200,000 adults in the US for a research app that answers health questions with OpenEvidence's cited medical AI, and shares no data with model training.

Apple has launched the Apple Health AI Usage Study in the United States. It is a virtual research study, with its own app, that will enroll up to 200,000 participants to examine how people use and trust conversational health AI. It is run with the clinical AI company OpenEvidence.
Who can join, and what the app does
The study is open to adults aged 21 and older in the US. Apple says participants need an iPhone 15 Pro or later with Apple Intelligence enabled, running iOS 26 or later. People enroll by downloading the study app. Apple does not mention a fee or a compensation scheme in the announcement.

Inside the app, participants can ask everyday health questions and get answers in plain language. Each answer links to peer-reviewed clinical citations. OpenEvidence supplies the medical AI, and Apple designed the research experience around it. According to Apple, OpenEvidence draws on sources such as The New England Journal of Medicine, the Journal of the American Medical Association and Nature.
Participants can also choose to share HealthKit data types, including resting heart rate, activity, sleep or lab results. Apple says the aim is to see how personal data changes the relevance and clarity of answers. A further strand of the study looks at whether health AI can help people prepare for visits with their own doctors. The company’s screenshots show a visit-prep summary and a survey that asks users to rate the quality of a response.
The study’s stated focus is what drives trust, the quality of responses, comprehension and everyday health actions. Apple says the results are meant to inform health AI design for itself and for the wider industry.
Privacy promises and their limits
Apple states that study data will not be used to train AI models. It also says participants control whether they share any HealthKit metrics or survey responses. The app includes guardrails meant to detect certain concerning questions and suggest following up with a provider or other resources. Apple adds that the app is not meant to diagnose or treat any condition or to replace medical advice.
The company said it has put security measures in place for the study data it maintains. The announcement does not say how long data will be kept, who outside Apple can see results, or whether OpenEvidence receives participants’ questions or HealthKit data. It also does not name an ethics board, a publication plan or a timeline for results. Those claims come from Apple and have not been independently verified.
Apple itself acknowledges the core risk. In its words, consumer health tools built on LLMs may sometimes give responses that are incomplete or unclear, which is why it says research is needed.
“Health is deeply personal, and conducting a dedicated research study is the most thoughtful way to understand how people engage with, comprehend, and evaluate health AI in everyday life.”
That is Sumbul Desai, M.D., Apple’s vice president of Health and Fitness. Robert M. Wachter, M.D., chair of the Department of Medicine at UCSF, is also quoted in the release. He argues that patients and clinicians use AI tools built and tested in separate worlds.
A cautious route into a crowded field
Apple is entering a market that already has large players. OpenAI launched ChatGPT Health on January 7, 2026. According to coverage of that launch, OpenAI said more than 230 million people ask health or wellness questions on ChatGPT each week. Its product lets users connect patient portals, Apple Health and wellness apps, though it excluded the EEA, Switzerland and the UK at launch, per Tom’s Guide and other reports.
Apple’s choice of partner is telling. OpenEvidence raised $250 million at a $12 billion valuation in January 2026, according to its own announcement. Press reports say more than 40% of US physicians use it daily and that it is free for doctors, funded by industry partners. Those usage figures come from secondary coverage rather than an audited source. Apple is borrowing a tool doctors already know rather than building its own medical chatbot, and framing the effort as research instead of a product launch.
That framing is the strategic point. A study of this size can give Apple data on how people react to health answers, with consent, before it commits to a consumer feature. The announcement makes no promise of one. It describes the work as research that informs design, and says nothing about a future Apple Health assistant. Anything beyond that is inference.
For readers, the practical change is narrow. Eligible iPhone owners in the US can try a cited, medically grounded AI chat, and may be asked to share sensitive health data. Everyone else, including users outside the US, those under 21 and owners of older iPhones, has no access. The study app is a research instrument, not a replacement for a doctor, and Apple says so.
Related reading: our report on Apple HomeView and where Siri AI is allowed shows how Apple’s AI features still depend on region and hardware.
Sign Up for the Daily AI Pulse
One email a day. All the stories that matter.
By signing up, you agree to our Terms of Use and acknowledge the data practices in our Privacy Policy. You may unsubscribe at any time.
Alex CarterAI News & Big Tech CorrespondentAlex Carter writes for Aiholics, keeping readers updated on the fast-paced world of AI and Big Tech.



