Apple announced the Apple Health AI Usage Study on November 15, launching a dedicated research app in partnership with clinical AI company OpenEvidence. The study aims to understand how consumers interpret and act on conversational answers to everyday health questions rather than offering a new, generally available diagnostic service.
Apple plans to enroll up to 200,000 people in the United States aged 21 and older. The company says participants need an Apple Intelligence-capable iPhone, specifically an iPhone 15 Pro or later, running iOS 26 or later.
What the study app actually does
Participants can ask the research app health-related questions and explore explanations supported by citations to medical literature. OpenEvidence contributes technology grounded in sources such as major peer-reviewed medical journals. Participants are asked to rate responses and can share feedback on whether answers are understandable and useful.
Users may choose to share some HealthKit information, including activity, resting heart rate, sleep or laboratory results, to explore whether personal context changes the relevance of responses. Apple says that data-sharing decisions remain with each participant and that study data will not be used to train AI models.
Research findings versus medical advice
Apple explicitly states that the study app is not intended to diagnose or treat a medical condition or replace clinicians. That matters because a medical answer that cites literature may still misunderstand a user’s situation, omit vital details or require a professional examination.
The research focuses not just on factual quality but also on trust and comprehension: whether people recognize uncertainty, understand the response and use it to prepare for a doctor’s appointment. Apple has described safety guardrails intended to flag certain concerning questions and direct participants to appropriate care.
In a broader AI sector increasingly interested in specialized scientific tools, Optimisus recently examined Google DeepMind’s AlphaGenome 2 research model. Apple’s study differs by examining how ordinary people use clinical information rather than making claims about discovering genetic mechanisms.
A study opening is not evidence that Apple plans to integrate the same service into Siri or the standard Health app. Publication of study findings, independent assessment and possible future product announcements remain separate milestones.
Source: Apple Newsroom, November 15, 2026.

