Apple is doing what Apple does: showing up late to an AI category and wrapping the arrival in a clinical trial. The company just launched a study app that lets people ask health questions and get answers tied to medical literature, with up to 200,000 participants over more than 12 months.
The partner is OpenEvidence, a healthcare-focused AI company. Answers draw on sources like The New England Journal of Medicine, the Journal of the American Medical Association, and Nature. That is a different pitch from ChatGPT or Claude, which answer health questions from broad training data and then add specialized features on top. Apple is betting that citation-backed, simplified answers to everyday medical questions are what people want and can understand.
Eligibility is narrow. Participants must be U.S. residents over 21, with an iPhone 15 or later, iOS 26 or later, and Apple Intelligence turned on. Users can rate answers, leave feedback, and complete monthly surveys. They can also share HealthKit data, such as resting heart rate, sleep, activity, or lab results, to get more personalized responses. The app can help prepare for a doctor’s visit and will point users to specific providers when a question sounds concerning.
Privacy is where Apple is leaning hardest, as expected. Data is anonymized before going to a secure third-party cloud provider, it won’t be used to train AI models, and OpenEvidence deletes query and response data after 24 hours. That matters because the HealthKit link is the real differentiator. OpenAI and Anthropic can offer health features, but Apple sits on years of passive sensor data and lab records. Whoever connects that data to a trustworthy model has an edge the others can’t easily copy.
The scale of demand explains the interest. OpenAI said earlier this year that users were sending over 300 million health-related queries each week. People are already using general chatbots for symptoms, lab results, and medication questions, with all the hallucination risk that implies. A study that measures how people comprehend and evaluate clinical answers is useful, and it gives Apple evidence if regulators or doctors ask hard questions later.
Sumbul Desai, Apple’s vice president of Health and Fitness, framed it as scientific rigor first, saying a dedicated research study is the most thoughtful way to understand how people engage with health AI. That’s careful language, and it’s also a hedge. Apple has not said whether this becomes part of the Health app or Siri. It is not clear that it will.
So does it matter? Somewhat. The app itself is a research tool, not a product, and a 12-month timeline means no shipping feature soon. Competitors will keep moving faster. But the combination of curated medical sources, on-device health data, and a privacy-first pitch is the most credible path to a health assistant that doctors might tolerate. If the results are good, expect Apple to fold this into Siri and Health, and expect OpenEvidence to benefit from the exposure. If they aren’t, Apple loses little and learns a lot, which is the advantage of being the cautious one.



