Apple and clinical AI platform OpenEvidence want up to 200,000 US participants to ask health questions and grade the answers
Apple wants to know whether AI health answers are actually any good. Rather than guess, it is asking up to 200,000 people to grade them.
The company announced the Apple Health AI Usage Study on November 15, 2026, in partnership with OpenEvidence, a clinical AI platform built for healthcare professionals. Participants will use a dedicated iOS app to ask medical questions, then score the responses on accuracy, detail and relevance.
How the study works
The study is virtual and limited to the US. Eligible participants must be 21 or older, and enrollment opened the same day as the announcement, with the app already available to download.
Once inside the app, users type in health questions the way they might ask a friend who went to medical school. The answers come from OpenEvidence, which grounds its responses in peer-reviewed literature, including The New England Journal of Medicine, JAMA and Nature.
After each answer, participants rate what they received. The research focuses on four areas: user trust, response quality, comprehension, and the effect on health decision-making.
Participants control their survey responses, and privacy protections apply throughout the study.
A shift from sensors to conversations
The study departs from Apple’s earlier health research, which centered on data collected by its devices. The Apple Health Study, which began in February 2025, is one example of that device-first approach.
Who OpenEvidence is
OpenEvidence may not be a household name, but it has become a fixture in many clinicians’ workflows. Its user base among monthly active US clinicians grew from 355 in January 2024 to 703,265 by June 2026.
The platform has also handled 271 million queries over 30 months.
OpenEvidence reached a valuation of approximately $15 billion after a $250 million funding round in September 2026.
The company has also integrated with major US health systems, pushing clinical AI deeper into everyday hospital operations.
Why the timing matters
The launch comes as scrutiny of AI in healthcare intensifies. In mid-October 2026, JAMA published a comprehensive analysis of how clinicians use these tools, adding fuel to an already active debate.
What this means
For Apple, the study offers something valuable: large-scale, structured feedback on how people engage with health AI before committing to a consumer product. With up to 200,000 participants, the dataset could give the company a clearer picture of where these tools earn trust and where they lose it.
The stated goal is to improve conversational health AI design for Apple and for the wider healthcare industry.
The study’s focus on comprehension suggests both companies are aware that translating medical literature for laypeople without distorting it is a hard design problem.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.

1 hour ago
7








English (US) ·