Skip to content
Featured on Startup Fame

Apple opens 200,000-user study to test AI health answers

Apple just opened the doors to a year-long study that will test AI-powered health answers for 200,000 US users. The app, built with health AI firm OpenEvidence, pulls its answers from peer-reviewed journals—not the unverified web data that fuels most chatbots. Participants must be 21 or older, own an iPhone 15 or later, and enable Apple Intelligence to enroll.

Editor, Lazyfounder

Published 4 min read
Apple opens 200,000-user study to test AI health answers
Image: Mint (Technology) via source

30 SEC SUMMARY

  • Apple teamed up with health AI firm OpenEvidence to launch a 12-month study testing AI-driven health answers for 200,000 US users.
  • The study app pulls answers from peer-reviewed journals like The New England Journal of Medicine, JAMA, and Nature—not unverified data.
  • Participants must be 21+, own an iPhone 15 or later, and enable Apple Intelligence to join.
  • Users can share health metrics via HealthKit for personalized responses but won’t have their data used to train AI models.
  • OpenEvidence deletes all user queries and responses within 24 hours to protect privacy.

Apple kicks off 200,000-user AI health study

Apple rolled out a research app this week to test how artificial intelligence can answer health questions—without repeating the mistakes of general chatbots. The app, built with health AI firm OpenEvidence, aims to sign up 200,000 US participants over the next 12 months.

Who can join and what participants do

To join, users must be 21 or older, own an iPhone 15 or later, and run iOS 26 with Apple Intelligence switched on. Once enrolled, they can ask everyday health questions, rate the AI-generated answers, and submit feedback through monthly surveys.

Where answers come from: peer-reviewed clinical journals

Instead of scraping the web like most chatbots, the app pulls answers from The New England Journal of Medicine, Journal of the American Medical Association (JAMA), and Nature. Apple says this system simplifies complex medical language so users can quickly understand health topics. If a question flags a serious condition, the app directs users to consult a healthcare provider.

Privacy: HealthKit data stays separate from AI training

Participants can optionally share data like resting heart rate, activity levels, sleep patterns, and lab results through Apple’s HealthKit. Those metrics let the AI tailor responses, but Apple promises the data stays de-identified and is never fed into any AI training pipeline. OpenEvidence, the study partner, will wipe all user queries and responses within 24 hours.

What’s next: no promises, but Health app integration looms

Apple has not said whether this technology will land in the Health app or Siri. For now, the company calls the study a way to learn how people interact with AI in healthcare while keeping clinical standards and privacy intact.

Why this matters: AI’s credibility crisis in healthcare

General-purpose chatbots like those from OpenAI and Anthropic have faced criticism for giving inaccurate or misleading health advice. Regulators and users have grown skeptical after high-profile failures, including Apple’s own delayed Siri AI upgrade and product flops like the Humane AI Pin. Apple’s study sidesteps those pitfalls by tying every answer to peer-reviewed sources.

How Apple’s approach stacks up against rivals

Most AI companies train their models on vast but unverified datasets, which can lead to errors or hallucinations. Apple’s choice to ground answers in NEJM, JAMA, and Nature sets a stricter standard. Competitors like OpenAI and Anthropic have added health features, but none have matched Apple’s scale or clinical rigor. The 200,000-user study could become the largest real-world test of clinically validated AI health advice.

What this means

Lazyfounder analysis — our interpretation, not reported fact.

Apple’s year-long study is testing whether clinical-grade AI answers can earn user trust—something general chatbots have failed to do. If the feedback from 200,000 participants shows people act on these answers, Apple could embed a similar tool in its Health app or Siri without the privacy and accuracy risks that sank earlier rivals. For health-tech founders, the study’s scale and rigor create a benchmark to measure their own AI tools against before Apple turns this research into a consumer product.

Key takeaways

  • Apple is anchoring AI health answers in peer-reviewed clinical literature, sidestepping the accuracy pitfalls of general chatbots.
  • The 200,000-user sample will generate a dataset large enough to set a new industry standard for trustworthy AI health advice.
  • If successful, Apple could integrate this engine into its Health app or Siri, raising the compliance bar for every competitor.
  • Startups have 12 months to test their own AI health tools against Apple’s gold-standard dataset before the company potentially turns it into a product.

Sources

  1. Mint (Technology) · 2026-11-16
    Apple launches AI health study app in partnership with OpenEvidence: Who can participate and how it works

This story is an original summary drafted with AI by Lazyfounder from the reporting listed above and checked by automated validation. Facts are attributed to their original publishers; sections marked as analysis are Lazyfounder's. Where a source is in another language, facts were machine-translated and quotations are reported, not reproduced. Read the original coverage via the links, and see our AI policy and corrections policy.

About the author

Editor, Lazyfounder

Tarun Mottlia edits LazyFounders, covering Indian startups, funding rounds, AI and product launches. Every story on the site is AI-assisted and checked against its cited sources before publication.

More stories by Tarun Mottlia

Get the LazyFounder Brief

Startup, funding and AI news in a five-minute read. Join the early-access list.

Lazy Founder - Powered by Blogy.in

Contact us

Have a story tip, correction or partnership idea?

Write to us at tarun.kumar@blogy.in or talk to the founder directly. We read every message.

Contact us