Apple launches OpenEvidence health AI research study | Magica
Apple takes OpenEvidence's medical AI into a consumer research study
Editorial Team
••📖6 min read
Apple's new U.S. study will examine how up to 200,000 adults understand and use health answers from OpenEvidence, a clinician-focused AI company. Optional personal health data adds context, but competing chatbots already offer data connections, and the study has yet to demonstrate better health decisions.
Apple is recruiting up to 200,000 U.S. adults to study conversational health AI, with optional access to personal health metrics.
The collaboration extends OpenEvidence's clinician-focused technology to consumers; January reporting described advertising to clinicians as its primary business model.
Personal-data connections already feature in competing health AI offerings. Apple's study has yet to show whether its approach improves decisions.
Apple, the iPhone maker, launched the Apple Health AI Usage Study with OpenEvidence, a company that makes a chatbot for clinicians to search medical evidence, on November 15. The experimental U.S. app lets eligible adults ask everyday health questions and optionally connect personal health data. Apple says the research will examine response quality, comprehension, trust and everyday health actions in its launch announcement.
The partnership brings technology developed for medical professionals to people who must describe their own concerns, interpret an answer and decide what to do next. That distinction matters: a separate randomized study found that access to capable language models did not improve users' choices about seeking care. Clinical citations alone cannot establish that a consumer tool works.
Apple's promotional screenshots show the study introduction and its conversational interface, developed with OpenEvidence. Image: Apple. Source: Apple.
A research app with optional personal context
Apple designed the research experience; OpenEvidence developed the medical AI. Apple says participants can read answers in approachable language and follow linked, peer-reviewed clinical citations, drawing on publications including The New England Journal of Medicine, JAMA and Nature. Participants can choose to share resting heart rate, activity, sleep or laboratory results through HealthKit, Apple's system for sharing health data with apps.
Apple wants to learn whether that personal context changes an answer's relevance and the user's understanding of it. The study will also explore whether the app helps participants prepare for appointments with their own doctors.
The FBI warns that ARTEX can accelerate intrusions, with some reaching privileged access in under three hours. Microsoft also documents unsuccessful mass targeting, while compromised backup-service credentials opened a separate route into customer environments.
Editorial Team
Enrollment is open to adults age 21 and older in the United States, with a ceiling of 200,000 participants. Apple specifies an Apple Intelligence-enabled iPhone—iPhone 15 Pro or later—running iOS 26 or later. Those requirements define the population the study can reach; even a large enrollment would leave younger people and people without eligible devices outside its scope.
Apple plans to run the study for more than 12 months, with users rating answers, providing feedback and completing monthly surveys, according to a report on the launch. The participant ceiling is a recruitment target, not an enrollment result.
Why these two companies
Apple brings experience recruiting people into virtual health research. In February 2025, it launched the Apple Health Study with Brigham and Women's Hospital, a research hospital and Harvard Medical School teaching affiliate. Apple said then that its earlier women's health, hearing, and heart and movement studies had more than 350,000 U.S. participants combined. It also credited research insights with helping develop products including the Apple Watch's Vitals app and the iPhone's Walking Steadiness feature. That history gives the new study a product-development precedent, although it does not establish the AI app's effectiveness.
OpenEvidence, founded in 2022, announced a $250 million funding round in January 2026 at a $12 billion valuation, according to a funding report. At that time, its chatbot was free to clinicians with a national provider identifier, and its primary business model was advertising shown to clinicians.
The financing makes the study a consumer experiment for a heavily funded commercial business. It does not establish that the app will carry advertising or adopt the clinician service's business model. The launch announcement does not disclose the partnership's financial terms, deployment costs or a future consumer pricing plan.
Apple said it wanted a clinically grounded partner when explaining the choice of OpenEvidence at the HLTH conference in Las Vegas, according to the launch report.
The market already includes health-focused chatbot offerings from OpenAI and Anthropic, as the launch coverage notes. Personal-data integration is also an existing approach: OpenAI's January announcement of ChatGPT Health, its dedicated health and wellness experience, described connecting medical records and wellness apps to inform conversations. Apple's contribution therefore depends on evidence about how consumers use its particular combination of medical answers, personal context and research design.
There is a reason to test people using the system as well as the system's answers. A randomized study published in February 2026 in Nature Medicine tested 1,298 UK adults in August–October 2024. Participants assessed up to two of ten fictional medical scenarios with GPT-4o, Llama 3, Command R+ or their usual resources. Controls mostly used internet search or their own knowledge.
The control group had 1.76 times the odds of identifying at least one relevant condition compared with the AI-assisted groups combined, with a 95% confidence interval of 1.45–2.13. Choices about the appropriate level of care, from self-care to an ambulance, did not differ significantly between each model-assisted group and controls. Researchers observed users omitting important information and failing to carry correct model suggestions into their answers.
This was a test of older models and fictional scenarios, not the Apple–OpenEvidence app or actual patient outcomes. It cannot predict the new study's results, but it shows why medical knowledge benchmarks alone cannot establish consumer usefulness.
Apple itself warns that language-model responses can be incomplete or unclear. It says the app includes guardrails intended to detect certain concerning questions and suggest contacting a provider or other resources. The app is not intended to diagnose or treat conditions or replace professional advice.
Privacy is another part of the experiment's credibility. Apple says participants control whether to share HealthKit metrics and survey responses, and study data will not train AI models. The technology launch report additionally attributes to Apple that data is anonymized before transmission to a secure third-party cloud provider and that OpenEvidence deletes query and response data after 24 hours. That reported deletion period concerns OpenEvidence's query and response data; it does not describe the lifetime of every study record.
Apple's study app asks participants to rate responses and identify useful qualities. These ratings are part of the research, rather than evidence of improved health decisions. Image: Apple. Source: Apple.
What the study still needs to establish
The next substantive milestone is evidence from participants. The announcement offers no results demonstrating improved comprehension, safer decisions or better preparation for appointments. It also does not specify a comparison group or a timetable for publishing findings.
Whether the experiment becomes a feature in Apple's Health app or Siri remains unclear, according to the technology launch report. A useful research outcome would distinguish answers people like and trust from answers they understand and act on appropriately—and explain how far findings from eligible adult iPhone users can travel beyond that group.
Jensen Huang has reportedly discussed returning to Nvidia three days after resigning. Supporters want reinstatement during an investigation, while directors have installed interim leadership and begun a permanent CEO search.
House China committee chairman John Moolenaar demanded answers from Commerce Secretary Howard Lutnick over delayed restrictions on Nvidia customer Megaspeed. The dispute concerns enforcement, overseas access to advanced chips and commercial influence as President Donald Trump prepares to visit China.
Apple reportedly still aims to unveil iPhone-linked glasses at WWDC 2027 and ship before year-end. Siri control of existing apps is the software bet, but delays cloud the schedule and Meta already has an established eyewear lineup.
Treasury Secretary Scott Bessent urged Nvidia to recruit an outside CEO and explain within 10 days whether suspect chip sales inflated earnings. The board is considering internal and external candidates, while export enforcement also faces scrutiny.
Congressional scrutiny and company investigations are testing what Nvidia knew about Megaspeed, a Singapore cloud operator whose Malaysian subsidiary bought nearly $2 billion of its technology. Megaspeed denies wrongdoing; the available reporting has not conclusively traced suspected chip diversion to China.
Commerce Secretary Howard Lutnick is backing narrower restrictions on Chinese access to overseas AI computing, according to a report examining his former firm’s financing relationships. Bitdeer filings establish Cantor’s role and the scale of its client’s planned expansion, but do not establish improper influence on policy.
Microsoft said services were fully restored on Saturday after a global Microsoft 365 outage disrupted email, healthcare communications and public services. It blamed an earlier network capacity upgrade exposed by a later change, with a full technical review still to come.
Cursor says a key exposed by Fireworks led an attacker to unredacted credentials in historical coding-agent records and then its extension cache. Up to 5,000 users potentially received malicious Claude Code builds. Cursor has moved inference off Fireworks; successful endpoint compromises remain uncounted.
Police are investigating a top tax-accountant exam scorer's admission of AI-assisted cheating, while the government reviews national qualification tests held in 2026 and prepares stronger controls.
An early tally puts Japanese listed companies’ first-half net profit growth at 43.2%, led by manufacturers. Filings show AI component demand alongside currency benefits and non-AI earnings gains, while procurement delays and canceled infrastructure reservations expose the costs of expansion.
TSMC plans to back promising chip startups through a new investment arm, alongside separate commitments to OpenAI and Anthropic. The strategy reaches toward future customers and the AI developers shaping chip demand, but the investment rights remain unclear.
Nidus has raised $30 million across pre-seed and seed rounds to automate wire processing in existing U.S. defense factories. Its founders bring manufacturing and government-deployment experience, but customer access and proprietary data still need to translate into measurable production gains.
Nebuly is funding US and European expansion for its AI business-value platform. Adjustable assumptions and private deployment are part of its pitch, but rivals offer similar measurement tools and verified customer returns remain undisclosed.
Han Won-seok says he used a miniature camera, generative AI and a smartphone in tax accountant and CPA exams. His confession puts scrutiny on exam security after proctors reported no suspicious behavior.
Reports linked to AI and child sexual exploitation rose from about 4,700 in 2023 to more than 400,000 in 2025, Childlight finds. The increase exposes growing threats and weaknesses in reporting systems, but does not measure unique victims or confirmed AI-generated images.
Megaspeed says Alice Huang left on August 30, 2025, and has retained outside lawyers for an internal compliance investigation. Singapore’s inquiry began months before that departure; authorities later found insufficient evidence that Huang had committed a crime, while the investigation into the company continues.
HP’s US listing puts a $132,000 starting price on its ZGX Fury shared AI system. Its 748 GB of coherent memory supports a local-computing pitch, but competing GB300 machines, dedicated power requirements and workload-specific operating costs shape the buying decision.
A $1.5 trillion valuation underpins the estimate of 4,250 Anthropic employee millionaires. The offering remains unpriced, selling restrictions limit liquidity, and the latest filing shows computing costs complicating the company’s rapid growth.
Supermicro says Rubin racks are shipping and Nvidia claims production at scale, while supplier reporting points to larger deliveries in 2027. Memory constraints and differing forecast scopes leave the pace of the transition unresolved.