Apple is recruiting up to 200,000 Americans for a research project that will examine whether medically grounded AI can help people understand health information, make informed decisions and prepare for doctor visits.
Apple launched a major research study on November 15, 2026, to investigate how people use artificial intelligence to find and understand health information. The project will enroll up to 200,000 adults in the United States and examine whether AI-generated medical answers can help people make better-informed decisions about their health.
The Apple Health AI Usage Study is being conducted in partnership with OpenEvidence, a company that develops AI-powered clinical information tools. Participants will use a dedicated iPhone application to ask health questions, receive answers supported by medical research and evaluate the information they receive.
Unlike a traditional clinical trial testing a medication or treatment, this project focuses on how people interact with health information. Researchers want to understand whether users trust AI responses, comprehend medical explanations and take appropriate action after receiving information.
The findings could help shape the future of consumer health technology, although Apple has not announced plans to introduce the study’s features into its regular Health app.
How Apple’s New Health AI Study Works
Participants will interact with an AI chatbot through an application called Apple Health AI Study. The app is separate from Apple’s existing Health and Research applications.
Users can ask questions about common health concerns, symptoms, recovery from medical procedures, exercise and other health-related subjects. The system is designed to explain medical information in language that people without clinical training can understand.
One important feature is the inclusion of citations linking answers to peer-reviewed medical research. Instead of simply providing a response, the application allows users to examine the scientific sources behind the information.
Participants can evaluate the answers and indicate whether they were understandable, relevant and useful. Researchers will then examine how people respond to medical information and whether the AI helps them identify appropriate next steps.
Apple also wants to study how people use AI before visiting healthcare professionals. The app includes features intended to help patients prepare questions and organize information for appointments.
Researchers will collect feedback after medical visits to determine whether using the chatbot helped participants feel better prepared or communicate more effectively with their doctors.
OpenEvidence Provides the Medical Research
Apple selected OpenEvidence to provide the underlying clinical AI technology for the study.
OpenEvidence develops tools that help healthcare professionals search and interpret medical research. Its platform draws on published studies, clinical guidelines and other established medical resources.
The company has partnerships involving major medical publications and research organizations, including The New England Journal of Medicine, JAMA, Nature, Cochrane and the National Comprehensive Cancer Network.
These sources contain research on medical treatments, diseases, prevention and other health subjects. The partnership gives Apple’s research application access to information from established clinical publications.
However, using reputable medical sources does not guarantee that every AI-generated response will be accurate or appropriate for an individual patient. AI systems can misinterpret information, overlook important details or provide explanations that users misunderstand.
Those limitations are among the reasons Apple is conducting the research.
The company wants to determine whether medical citations and clearer language improve understanding. Researchers will also examine whether users place appropriate confidence in answers rather than trusting them simply because they appear authoritative.
Personal Health Data Could Make Answers More Relevant
One of the study’s most significant features is the option to connect personal information already collected through Apple’s Health app.
Participants may choose to share measurements such as resting heart rate, physical activity, sleep patterns and laboratory test results. This information could allow the AI to provide answers that reflect a person’s own health measurements.
For example, someone asking about resting heart rate could receive an explanation that considers their recorded measurements rather than relying entirely on general population information.
Users who have imported clinical records into Apple’s Health app may also choose to share certain records with the research application.
The ability to combine personal measurements with medical literature is an important subject of the study. Researchers want to determine whether this additional context makes answers more useful and easier to understand.
Sharing health information is optional. Participants can use the study application without granting access to their HealthKit measurements.
The research does not establish whether personalized AI responses improve health outcomes. That is one of the distinctions between testing how people understand information and demonstrating that a technology improves medical care.
Apple Builds Privacy Protections Into the Research
Privacy is a major consideration because the study involves potentially sensitive medical questions and personal health information.
According to Apple, questions submitted through the application first undergo processing on the participant’s iPhone to remove identifying information.
The questions then pass through a secure environment managed by a third-party research support contractor. There, they are separated from individual participant identifiers and screened again before being sent to OpenEvidence for processing.
Apple says the study data will not be used to train AI models. OpenEvidence will also delete question-and-response information within 24 hours, according to the company’s statements to reporters.
These protections are intended to reduce the amount of identifiable information reaching the AI provider.
However, removing personal identifiers does not eliminate every potential privacy risk. Health information can contain unusual details that may make individuals recognizable, particularly when combined with other information.
Participants will be asked to review the study’s consent materials before enrolling. Understanding what information is collected and how it is handled will be important for anyone considering participation.
The research is overseen by an independent review board, which evaluates ethical safeguards for studies involving human participants.
Researchers Want to Know Whether AI Changes Health Decisions
The study addresses a growing question in healthcare: Does having access to an AI-generated medical explanation actually help someone make a better decision?
A chatbot might provide a technically accurate description of a medical condition, but a patient could misunderstand its meaning or apply the information incorrectly.
For example, a person might interpret general information about a symptom as reassurance that medical attention is unnecessary. Another person might become alarmed by a list of possible conditions that have little relevance to their situation.
Dr. John Brownstein, a professor of pediatrics and biomedical informatics at Harvard Medical School, told ABC News that researchers need to examine whether people understand AI answers and use them appropriately.
The study will explore these questions through participant feedback, surveys and evaluations of how the information affects everyday decisions.
Apple has also included safety features intended to recognize certain concerning questions. When appropriate, the application may direct users toward healthcare professionals or other medical resources.
These safeguards do not mean the application can reliably detect every medical emergency or dangerous situation.
Apple emphasizes that the study application is not designed to diagnose diseases, recommend treatment or replace professional medical care.
Who Can Participate in the Apple Health AI Study?
Apple is accepting enrollment from eligible adults throughout the United States. The study has a maximum planned enrollment of 200,000 participants.
To qualify, volunteers must be at least 21 years old and have an Apple Intelligence-enabled iPhone running iOS 26 or later. Supported devices include the iPhone 15 Pro and newer compatible models.
Interested users can download the Apple Health AI Study app from the App Store and review the enrollment requirements.
Participants must complete an informed consent process and provide demographic information. Researchers will examine factors such as age, education and other characteristics to better understand differences in how people interpret AI-generated medical information.
The study includes an initial survey and monthly follow-up questionnaires. Participants may remain enrolled for up to 12 months.
Some surveys will focus on how individuals use the information they receive. Others will examine whether the chatbot helped them prepare for medical appointments or communicate with their healthcare providers.
Apple has not announced a specific date for publishing the research findings.
What the Study Could Mean for Future Healthcare Technology
The research comes as AI-powered health information becomes increasingly accessible to consumers.
People can already ask general-purpose chatbots about medical conditions, medications, test results and symptoms. However, these systems are not always designed specifically for clinical use.
Apple’s research focuses on whether AI technology supported by established medical literature can provide information that patients understand and use responsibly.
Sumbul Desai, Apple’s vice president of Health and Fitness, said the company believes responsible innovation requires scientific evaluation and a strong clinical foundation.
The project may also help researchers understand how AI tools can complement medical appointments rather than competing with healthcare professionals.
Robert Wachter, chair of the Department of Medicine at the University of California, San Francisco, emphasized the potential benefits of patients and clinicians using trustworthy information tools that support better communication.
The important unanswered question is whether these systems can consistently deliver understandable, reliable information without encouraging misplaced confidence.
Apple’s new study is intended to provide evidence about those real-world interactions. The company has not committed to releasing a consumer health chatbot based on the research, and no results are available yet.
For now, the Apple Health AI Usage Study offers researchers an opportunity to observe how thousands of people use medically grounded AI and whether those interactions help them navigate everyday health questions.
IMAGE: Source Black Apple Watch on my wrist #jcutrer Author Jonathan Cutrer CC2
