Apple has opened a virtual research study, run through a dedicated app built on OpenEvidence's clinical AI, to measure how consumers use conversational health chatbots. The companies announced it on stage at the HLTH conference in Las Vegas on 15 November 202612.
What the study covers
The study is open to up to 200,000 eligible adults aged 21 and over across the US, according to Apple's press release as reported by Healthcare Dive1. Participants use the app to ask questions about their health and look up medical information. They can choose to connect personal data, including resting heart rate, activity, sleep and lab results, so that answers reflect their own records12.
Apple designed the study, while the AI behind it comes from OpenEvidence, said Dr Sumbul Desai, Apple's vice president of Health and Fitness1. She set out the research questions directly: “We’re studying what builds user trust, how people comprehend responses, the quality of responses, and we wanted to work with a clinically grounded partner, and that is OpenEvidence”12. The study will also test whether a chatbot can change behaviour1.
Several design choices respond to known weaknesses in consumer health AI. Apple said the app will spot worrying questions and point users to an outside provider or other resources1. Data sharing is opt-in, and the company said study data will not be used to train AI models1. Apple also wants to learn how to present complex medical information more clearly, since products built on large language models often give incomplete or confusing answers1.
The app is also pitched as a tool to prepare for appointments. OpenEvidence chief executive Daniel Nadler said consultations frequently last only 10 or 15 minutes, so questions can be missed, and the app could help patients bring their data and concerns to the visit1.
What it means for the market
The pairing joins two strong distribution channels. OpenEvidence told NBC News earlier this year that its AI medical search tool is used by almost two-thirds of US physicians1. Nadler framed the logic plainly, calling OpenEvidence “the most trusted and loved brand product among doctors” and Apple the equivalent among consumers2. A consumer product carrying a clinician-facing brand is a different proposition from general-purpose chatbots, and health systems will need to consider how patients arrive at appointments having already consulted it.
Healthcare Dive notes that Apple could use the findings to shape its own AI products for healthcare1. That places the study in a wider contest among wearables makers. Google and Oura are adding AI coaches that turn raw device metrics into guidance1. Demand exists: roughly one-third of US adults have used an AI chatbot to discuss their health, according to KFF figures cited by Healthcare Dive1.
For regulators and clinicians the study could produce evidence on trust and comprehension that consumer health AI has so far lacked. STAT reported that the launch hints at possible new directions for both companies2.
Caveats
This is the start of a study, not a result. No findings have been published, and the cohort is limited to US adults who opt in through the app1. Concerns about hallucinations, incomplete answers and missing safety guardrails remain for specialised language models, even as their answers to clinical questions improve1. The study follows Apple's earlier research on mobility, heart health, hearing and fertility, and the Apple Health Study it began last year with Brigham and Women's Hospital1.
Sources
- Apple, OpenEvidence launch conversational health AI study and app (opens in a new tab)Healthcare Dive · healthcaredive.com
- Apple and OpenEvidence launch experimental health chat app (opens in a new tab)STAT · statnews.com
This briefing summarises publicly available research and reporting for information only. It is not medical, investment or legal advice. Follow the sources listed for the primary research and reporting.