Quick Answer
Apple and OpenEvidence have launched a study app that gives U.S. users access to clinically grounded health AI so researchers can examine how people ask questions, understand answers and rate the results. The study is planned for up to 200,000 participants over more than 12 months.
What Is Apple’s Health AI Study App?
The Apple Health AI Usage Study is a research app developed with OpenEvidence, a clinical AI platform. Participants can ask everyday medical questions and receive simplified answers based on clinical material from The New England Journal of Medicine, Journal of the American Medical Association and Nature.
Apple is inviting U.S. users who are at least 21 years old, use an iPhone 15 or later, run iOS 26 or later, and have Apple Intelligence enabled. The app is available for download in the United States.
Researchers plan to include up to 200,000 participants for more than 12 months. The study will examine how people use the app, whether they comprehend its responses, how they evaluate the information and whether they apply it to everyday health decisions.
How Does the App Answer Health Questions?
The app provides answers grounded in clinical data rather than serving as a general-purpose chatbot. Apple said the study app is intended to simplify common medical questions while linking participants to clinical publications supporting the responses.
Participants can rate each answer and submit feedback about the response. They also receive monthly surveys, giving researchers repeated information about their experiences rather than relying only on a single interaction.
The research will consider questions about preparing for doctor visits. Apple also says the app can direct users toward specific providers for concerns that may require clinical attention.
Can Participants Share HealthKit Data?
Participants can choose to share selected HealthKit information, including resting heart rate, physical activity, sleep data and laboratory results. The study will assess whether this personal information produces more relevant and understandable suggestions.
Sharing HealthKit information is optional. Participants retain control over whether they provide those data and their survey responses, according to Storyboard18’s account of the program.
How Does Apple Handle Study Data?
Apple says participant data will be anonymized before it is sent to a secure third-party cloud provider. The data will not be used to train AI models.
OpenEvidence deletes query and response data after 24 hours, according to Apple. The app also contains safety mechanisms intended to recognize certain concerning questions and encourage users to consult healthcare providers or seek appropriate resources.
“Health is deeply personal. We believe conducting a dedicated research study is the most thoughtful way to understand how people engage with, comprehend, and evaluate health AI in everyday life,” said Sumbul Desai, MD, Apple’s vice president of Health and Fitness.
Desai said the findings would help inform the development of trusted health tools as the technology evolves.
What Does the Research Add to Health AI Evaluation?
Apple’s study is examining a consumer-facing use of health AI: people ask questions, review simplified clinical answers, rate them and optionally connect those answers with data from their own HealthKit records. The enrollment target of up to 200,000 participants and planned duration of more than 12 months make the project a sustained research effort rather than a short product survey.

A separate randomized trial led by researchers from the UW School of Medicine and Public Health and UW Health evaluated a different application of health AI. Published in NEJM AI, that study found ambient AI notetaking reduced documentation time by 30 minutes per day per provider and correlated with a clinically meaningful reduction in practitioner burnout scores.
The ambient AI system passively listened to doctor-patient visits, drafted notes and required a healthcare professional to review the generated notes for quality and accuracy. Its trial ran from August 2024 through March 2025. It does not study Apple’s consumer app, but it adds a research example focused on documentation and practitioner well-being rather than patient use of a health question-and-answer tool.
The World Health Organization’s S.A.R.A.H. project offers another research example involving generative AI in public health. The prototype operated in eight languages, offered round-the-clock conversations about topics such as nutrition, stress reduction and quitting tobacco, and gathered data for iterative evaluation. It is not currently in active public deployment.
Key Considerations
- Enrollment is limited to U.S. users who are at least 21, use an iPhone 15 or later, run iOS 26 or later, and have Apple Intelligence enabled.
- Participation in the study may last more than 12 months and includes monthly surveys, answer ratings and response feedback.
- HealthKit sharing is optional and can include resting heart rate, activity, sleep and laboratory results.
- Apple says the study data will be anonymized, will not be used to train AI models, and will be handled under a 24-hour deletion policy for query and response data at OpenEvidence.
- The study examines how users engage with health AI, but the source does not provide findings about the app’s safety or effectiveness.
What to Do Now
- Check whether you meet the study’s age, device, operating system and Apple Intelligence requirements.
- Download the app if you want to participate in the research.
- Read the app’s privacy terms before submitting questions or HealthKit information.
- Decide which HealthKit categories, if any, you want to share.
- Provide answer ratings and monthly survey feedback throughout the study.
Therabot Depression and Anxiety Trial
The Therabot Depression and Anxiety Trial in 2025 included 106 people from across the United States diagnosed with major depressive disorder, generalized anxiety disorder, or an eating disorder. Participants interacted with Therabot through a smartphone app by typing out responses to prompts about how they were feeling or initiating conversations when they needed to talk.

People diagnosed with depression experienced a 51% average reduction in symptoms, leading to clinically significant improvements in mood and overall well-being. Participants with generalized anxiety reported an average reduction in symptoms of 31%.
Therabot users showed a 19% average reduction in concerns about body image and weight, which significantly outpaced a control group that was also part of the trial. The researchers concluded that AI-powered therapy is still in critical need of clinician oversight, while it has the potential to provide real-time support for people who lack regular or immediate access to a mental-health professional.
FAQ
Who Can Join Apple’s Health AI Study?
The app is available to U.S. users who are at least 21 years old, use an iPhone 15 or later, run iOS 26 or later, and have Apple Intelligence enabled.
Who Is Developing the Health AI Study App?
Apple is developing the app in partnership with OpenEvidence, a healthcare-focused clinical AI company.
How Many People Can Participate in the Study?
Apple aims to include up to 200,000 participants and plans to run the study for more than 12 months.
What HealthKit Data Can Participants Share?
Participants may share resting heart rate, physical activity, sleep information and laboratory results. Participation in the study does not require sharing that information.
Will Apple Use the Study Data to Train AI Models?
Apple says the data will not be used to train AI models. Apple also says data will be anonymized before being sent to a secure third-party cloud provider, while OpenEvidence deletes query and response data after 24 hours.
Related Coverage
- Planetary Health Diet Linked to Better Diet Quality, Study Finds
- 10-Year Study Links Dietary Variety and Dental Health to Depressive Symptoms in Older Adults
