How AI is Changing Healthcare: Apple's New Research Study

Apple launched a Health AI Usage Study with OpenEvidence to test how people trust and understand AI health answers. Up to 200,000 US iPhone users can join. The app flags risky questions and urges doctor visits, though findings carry real participation limits.
How AI is Changing Healthcare: Apple's New Research Study
Written By:
Murali Teja
Reviewed By:
Achu Krishnan
Published on: 
Updated on: 

Overview

  • Apple launched the Apple Health AI Usage Study with OpenEvidence on November 15, 2026, open to eligible adults across the United States.

  • The research checks whether people trust AI-generated health answers, understand them, and use them before doctor visits.

  • The study app flags risky questions and points users toward a doctor, but its findings still carry limits tied to geography, devices, and who chooses to join.

A health-related query often starts with a chatbot now, not a doctor's visit. More people ask AI about symptoms and medical concerns every day. That raises a bigger question. Can they trust what these tools say? 

Apple wants an answer. Its new Health AI Usage Study, built with OpenEvidence, looks at how people use AI for health questions. It checks whether people understand the answers, and how much they trust them when making real health decisions.

How Apple and OpenEvidence are Joining Forces 

Apple designed the study app and the research experience around it. The project pairs medically grounded AI technology from OpenEvidence with that Apple-built experience. The answers draw on research published in sources such as The New England Journal of Medicine, the Journal of the American Medical Association, and Nature. 

The tech giant is testing how people respond to clinically grounded answers. It is not building a new medical AI model from the ground up.

Why Apple Wants to Study Health AI Now

People keep turning to chatbots for health advice despite repeated cases of inaccurate or risky answers. Apple wants real data instead of guesses about what happens when ordinary people lean on AI for health questions. 

Dr. Sumbul Desai, Apple's vice president of Health and Fitness, called health deeply personal and said a dedicated research study offers the most thoughtful way to understand how people engage with, comprehend, and evaluate health AI in daily life.

Who Can Join Apple's Study, and What it Leaves Out 

The study is open to up to 200,000 eligible participants across the United States, who get access to an iPhone app that answers everyday health questions using medical research and, optionally, their own Health app data. Although the study could include up to 200,000 participants, its eligibility requirements limit the pool. Participants need a compatible iPhone, which leaves out people on Android devices, older phones, or living outside the US.

Another limitation concerns the people who choose to participate. People willing to join an AI health study may already trust AI more than most. The results could skew more positive than the average person's actual experience.

What Apple Wants to Learn About AI and Health 

Apple wants to know whether people trust AI health information and whether it feels clear, reliable, and genuinely useful. Participants can choose to share data such as resting heart rate, activity, sleep, and lab results through HealthKit for more personalized responses. 

A single health reading may offer limited insight, while patterns across several measurements can provide more context. But a pattern across sleep, activity, and heart rate over several weeks can point toward something worth raising with a doctor.

Apple is also studying how AI can help people prepare for clinical appointments with their doctors. This is a practical gap worth closing. Many people walk into short appointments without a clear list of questions. An AI tool that organizes those concerns beforehand could make that time count for more.

How Apple Plans to Make Health AI Safer 

The study app watches for what Apple calls concerning questions and prompts the user to consult a doctor when one comes up. This guardrail sits at the center of the whole study. Even an AI tool drawing on clinical research could give a confident response to a question that should prompt an urgent medical consultation. The feature is Apple's attempt to catch that gap before it reaches someone unprepared.

Why Trust Does Not Always Mean Accuracy 

A confident-sounding answer is not the same as a correct one. Medical research does not always apply the same way to every person, and a chatbot rarely signals that nuance on its own. 

A tool that answers quickly and clearly can lead people to accept the answer without checking it against a professional. This gap remains one of the hardest problems in health AI. No amount of clinical sourcing removes it entirely.

Also Read: AI in Patient Care and Personalized Treatment Plans

How Apple Plans to Protect Health Data

Apple says study data will not be used to train any AI models. Separately, questions pass through privacy protections designed to strip identifying information before reaching OpenEvidence. 

Sharing health metrics stays optional throughout. These are two separate privacy commitments. They describe Apple's stated protections, but do not mean participation carries zero risk. 

Apple told ABC News it plans to share the research broadly, aiming to inform computer-assisted health communication beyond its own products. If that holds, the findings could shape the wider development of AI health tools, not just Apple's own.

Also Read: Wearable Tech Revolutionizing Remote Patient Monitoring: Enhancing Care with Real-Time Health Insights

Final Thoughts

The real test ahead has little to do with how well an AI model can answer a medical question. It comes down to whether people can tell a reliable answer from a risky one, and whether they still reach for a doctor at the right moment. 

Apple's study marks one step toward answering that. The much harder step is finding out whether its results hold up for anyone outside this one narrow group of US iPhone users.

You May Also Like: 

FAQs

1. What is the Apple Health AI Usage Study?
It is a research study from Apple, run with OpenEvidence, that lets US adults ask everyday health questions through a dedicated iPhone app. The app answers using clinical research instead of general internet content.

2. Who can join the study?
Up to 200,000 eligible adults across the United States can take part. Participants need a compatible iPhone, so people on Android devices or living outside the US cannot join.

3. Does the study use personal health data?
Sharing health data stays optional. Participants can choose to share information like heart rate, sleep, and lab results through HealthKit for more personalized answers, but the study still works without it.

4. Will Apple use this data to train AI models?
No. Apple says data collected through the study will not be used to train any AI models. Separate privacy protections also strip identifying details before questions reach OpenEvidence.

5. What happens if someone asks a risky or urgent health question?
The app watches for concerning questions and prompts the person to consult a doctor. This guardrail is meant to catch situations where a quick AI answer is not enough.

Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp