Apple has opened a US study of how people use conversational health AI, giving eligible volunteers an app built with OpenEvidence. Joining means participating in research about trust, understanding and use of AI answers. It is not evidence that the app improves an individual's health or can replace clinical advice.
Apple announced the Health AI Usage Study on November 15, 2026. The originating CNET coverage is dated November 16. The announcement describes the study's aims and access conditions; it does not report completed results.
What volunteers can contribute
Participants can ask health questions, inspect linked clinical sources and provide feedback about responses. Apple says optional HealthKit inputs, including activity, sleep, resting heart rate or lab results, let the study explore how personal context affects relevance and understanding.
That creates two things to judge separately: whether an explanation is understandable and whether it is appropriate for a person's particular situation. A citation can help someone inspect the basis of an answer, but it does not by itself resolve that second question. The App Store description explicitly says neither the app nor the study provides medical advice, diagnosis or treatment.
Preparing questions for a medical visit is one intended use. An answer that sounds fluent should still be checked with a clinician when it concerns a health decision; the study is investigating how people engage with the tool, not handing participants a finished clinical verdict.
Eligibility and permission are different checks
Apple lists up to 200,000 US participants aged 21 or older, using an Apple Intelligence-enabled iPhone—described as iPhone 15 Pro or later—with iOS 26 or later. The store's 18+ age rating is not the study's enrollment threshold. A newer model name alone should not substitute for checking Apple Intelligence support and the app's eligibility flow.
Apple says sharing HealthKit metrics or survey responses is optional and study data will not train AI models. That promise answers a specific use-of-data question; it does not mean no information is collected or retained.
The general Health Studies Privacy Policy linked by the app, updated in 2020, says each study's informed consent controls when terms conflict. It also warns that withdrawing may not delete previously collected information and that deleting an app does not itself withdraw a participant. Read this study's consent for its actual data recipients, retention and withdrawal steps rather than treating the general policy as the complete study agreement.
Decide on the research terms before adding more data
For a prospective volunteer, the practical order is eligibility, consent and then a deliberate choice about optional data. There is no evidence-based reason in this announcement to buy a phone solely to join.
This is a different problem from connecting Samsung Health data to hospital workflows: an AI research conversation does not establish a connection to a person's care team. Claims about better health outcomes will require results beyond the launch announcement.