Apple Tests AI Health Answers With Cited Research

The study app links each response to peer-reviewed clinical citations and is still a research effort, not a general product.

Updated on Nov. 16, 2026 in Artificial Intelligence

Isometric editorial illustration of a tall stack of closed, unlabeled journal volumes with one pulled slightly out, evoking cited medical research.
Apple launched a research study app that answers personal health questions with links to peer-reviewed clinical citations, open to U.S. adults 21 and older. AI Illustration. Upload story photo >

Live Poll

Would you trust an AI chatbot to answer your health questions?

Apple has launched the Apple Health AI Usage Study app, which answers personalized health questions using trusted medical research and links each answer to peer-reviewed clinical citations. The study is a research effort, not a general-availability product, and it is open to U.S. adults aged 21 and older with an iPhone 15 Pro or later running iOS 26 or later.

Why it matters

The study measures whether users trust, understand and find helpful AI-generated health answers, and how those answers affect conversations with doctors. Apple says the aim is to make AI a bridge to better care, arriving as research has questioned how accurate chatbot medical advice is in practice.

The app's answers draw on peer-reviewed clinical citations that users can view alongside each response. Questions are processed with privacy protections meant to strip identifying information before they reach OpenEvidence, which deletes question-and-response data within 24 hours and does not train its model on the queries.

The players

Apple

Apple is the consumer hardware and software company behind iPhone, iOS and the Health app, and it is now running research on AI health answers inside its own ecosystem.

OpenEvidence

OpenEvidence is an AI medical reference tool used by doctors, and its full access is currently restricted to health care professionals.

The details

Users ask the study app personalized health questions and can optionally share health app data such as resting heart rate, activity, sleep or blood test results so responses are tailored to them. Each answer comes with linked peer-reviewed clinical citations, so readers can check the underlying research. The app uses OpenEvidence, an AI tool that doctors already use, and full OpenEvidence access is currently limited to health care professionals.

Timeline

  1. February 2026: An Oxford study on the medical accuracy of AI chatbots was published.

  2. Sunday, November 15, 2026: Apple announced the Apple Health AI Usage Study app.

The Tech Race

Apple's study follows a February 2026 Oxford study that found AI chatbots identified conditions in clinical cases with 95% accuracy, while everyday users reached the right answer on dense guidance only 35% of the time. Apple's approach pairs a consumer app with a physician-facing tool and peer-reviewed citations, a different bet from chatbots that answer without a verifiable source.

U.S. adults aged 21 and older with an iPhone 15 Pro or later running iOS 26 or later can take part in the study now, and they can share health app data if they choose. Availability of the app beyond the study, and any pricing, has not been announced.

The takeaway

The key trajectory is whether vetted, citation-linked AI answers change how patients and doctors talk about health. Watch for Apple's announced plan to make the research broadly available, since its results on user trust and comprehension will show whether this approach narrows the gap between chatbot accuracy and everyday use.

Further reading

For more on how these systems are being built and tested, explore our Artificial Intelligence coverage.

Live Poll

Would you trust an AI chatbot to answer your health questions?