Quick Facts

• The FDA had authorized 1,451 AI-enabled medical devices by the end of 2025, with 1,104 (76%) in radiology alone.
• 81% of US physicians reported awareness or use of AI in their practice in 2026, up from 38% in 2023. (AMA).
• Severe-harm potential from AI clinical recommendations ranged from 2.9% (best specialized tools) to 24.6% (general-purpose models) in the NOHARM benchmark. Doing nothing carried 37% severe-harm potential in the same test.

This page collects verified, primary-sourced statistics on AI in healthcare, organized by category. Every number here is traced to its original source rather than repeated from a secondary aggregator. Where a study has been revised or updated, we cite the current version and note it. This page is updated periodically as new data becomes available.

MARKET SIZE AND GROWTH

• The global AI-in-healthcare market was valued at $36.7 billion in 2025 and is projected to reach $50.7 billion in 2026, reaching $505.6 billion by 2033, a 38.9% compound annual growth rate from 2026. Other research firms publish different forecasts. Source: Grand View Research, June 2026 market report.
• North America held 54% of global AI-in-healthcare market revenue in 2025, the largest of any region. (Grand View Research).

FDA Regulatory Data on AI-Enabled Devices

• 1,451 AI-enabled medical devices had been authorized by the FDA by the end of 2025, with 1,104 (76%) of those in radiology. Source: The Imaging Wire, analysis of the FDA’s AI-Enabled Medical Device List.
• In Q4 2025 alone, the FDA cleared 72 AI-enabled devices, 55 (76%) of which were radiology tools, confirming the specialty’s continued dominance in new clearances. Source: The Imaging Wire.
• In Q2 2026, the FDA authorized 86 AI/ML devices spanning cardiovascular, neurology, GI/urology, anesthesiology, orthopedics, dental, surgery, hematology, pathology, clinical chemistry, and microbiology, signaling clearance activity is beginning to spread beyond imaging. Source: Innolitics.
• As of the FDA list update covering September 2025, no authorized device was LLM-based. The tools doing the heaviest clinical lifting today are narrow classifiers, not conversational AI. Source: “From Disclosure to Self-Referential Opacity: Six Dimensions of Strain in Current AI Governance”.

Physician Adoption of AI

• 81% of physicians reported awareness or use of AI in their practice in 2026 (AMA’s wording), more than double the 38% in 2023. Source: American Medical Association, 2026 Physician Survey on Augmented Intelligence (nearly 1,700 physicians surveyed).
• The average physician now uses 2.3 distinct AI use cases, up from 1.1 in 2023, according to the same AMA survey.
• The most common physician uses of AI are summarizing medical research (39%) and generating discharge instructions, care plans, or progress notes (30%): administrative and research support, not autonomous diagnosis.
• 77% of physicians say AI improves their ability to care for patients, up from 65% in 2023. At the same time, 88% worry about skill loss from over-reliance on AI, and 85% want a direct say in how AI gets adopted at their institutions, per the same AMA survey.

GLOBAL AI ADOPTION: HOW THE US COMPARES

• Globally, 48% of clinicians reported using AI for work in 2025, nearly double the 26% reported the year before. Adoption was highest in China (71%) and lowest in the US (36%) and UK (34%) in this survey.
• Despite 97% of AI-using clinicians having tried a generalist tool like ChatGPT, only 16% use AI for actual clinical decision-making, underscoring a gap between broad experimentation and deep clinical integration. (Elsevier’s Clinician of the Future 2025).

Note: These numbers aren’t directly comparable to each other. Elsevier’s 48% measures clinicians worldwide who have used AI “for work” in any capacity. The AMA’s 81% figure above counts US physicians who reported any awareness or use of AI in their practice. The gap between the two numbers reflects a difference in what was asked as much as a real difference in US versus global adoption.

AI ADOPTION IN PRIMARY CARE

• In a Doximity survey of 3,151 US physicians across 15 specialties (fielded March to April 2025 and November 2025 to January 2026; 54% of all respondents used AI in clinical practice), internists (the specialty closest to primary care) reported 60% AI adoption, among the three highest specialties alongside neurology (64%) and gastroenterology (61%).
• Family medicine physicians weren’t the top adopters by raw percentage, but were the heaviest users: 88% of family-medicine physicians who use AI use it daily.
• The most common use cases among physicians overall were literature search (35%) and voice-based documentation tools like AI scribes (29%), both up sharply from early 2025. (Doximity 2026 State of AI in Medicine Report)

Hospital and Health System Adoption

• A national survey of 2,174 nonfederal acute care hospitals found 31.5% were already using generative AI integrated with their EHR by 2024, with 24.7% planning to adopt within a year and 43.7% delayed or uncertain. Source: JAMA Network Open.
• Adoption correlates strongly with resources: major teaching hospitals (53.9%) and system-affiliated hospitals (38.5%) were far more likely to be early adopters than independent hospitals (16.3%), per the same study.
• Among US healthcare leaders surveyed by McKinsey (providers, payers, and technology firms), generative AI implementation rose from 25% in late 2023 to 47% in 2024, reaching 50% by the end of 2025, with more than 80% of surveyed leaders reporting at least one deployed use case. Source: Becker’s Hospital Review, McKinsey survey.
• Epic told Fierce Healthcare in February 2026 that 85% of its customers are live with generative AI tools (Art, Emmie, Penny). This is a company-reported figure. 

AI Diagnostic Accuracy: What the Studies Show

• Diabetic retinopathy screening systems show pooled sensitivity around 93% and specificity around 90% across regulator-approved systems. Source: npj Digital Medicine meta-analysis (82 studies, 887,244 examinations).
• A foundation-model-powered body CT triage tool covering 14 conditions posted 97% mean sensitivity and 98% mean specificity in its pivotal study (Aidoc, January 2026 FDA clearance; company-reported). Source: Aidoc.
• In a prospective, multicenter study of intracranial hemorrhage detection across 67 organizations analyzing 3,409 brain CT studies, radiologists using AI as an assistive tool significantly outperformed standalone AI: sensitivity 98.91% (AI-assisted) versus 95.91% (standalone); specificity 99.83% versus 87.35%. Source: Journal of Clinical Medicine.
• Fracture detection tools show pooled sensitivity around 90-92% and specificity around 91% across dozens of imaging studies. Source: “Artificial Intelligence in Fracture Diagnosis on Radiographs: Evidence, Pitfalls, and Pathways for Clinical Integration”.
• Using the New England Journal of Medicine’s archive of especially difficult diagnostic cases, one model’s first-guess diagnosis was correct in about 52% of 143 hard cases, with the correct answer appearing in its differential in about 78% of cases. On a head-to-head comparison of 70 overlapping cases, a newer model reached the exact or very close diagnosis in about 89% of cases, versus roughly 73% for an earlier model. Source: “Superhuman performance of a large language model on the reasoning tasks of a physician”.
• General-purpose generative AI averaged only 52.1% accuracy across 83 studies on open-ended diagnosis (95% CI: 47.0-57.1%), close to a non-expert clinician’s performance and well behind specialty-trained diagnostic models. Source: npj Digital Medicine.

Safety and Risk Data

• In the NOHARM benchmark (Stanford/Harvard), testing 20 generalist LLMs and 4 specialized clinical AI tools against 1,100 real case-based tasks, severe-harm potential reached up to 24.6% for the models tested, with more than 80% of severe errors being errors of omission (recommending too little) rather than active harm. Source: “First, do NOHARM”.
• Specialized clinical AI tools performed dramatically better than general-purpose models in that same benchmark: AMBOSS LiSA, Doximity Ask, OpenEvidence, and Glass Health with the lowest severe-harm rate at 2.9% (AMBOSS LiSA)
• A Mount Sinai study in Nature Medicine found ChatGPT Health under-triaged 52% of gold-standard emergencies (33 of 64 test cases) in a structured test of 60 clinical vignettes, often steering people away from urgent care when they needed it most. Source: Nature Medicine.
• A BMJ Open audit of five major AI chatbots (Gemini, DeepSeek, Meta AI, ChatGPT, Grok) found 49.6% of answers to 250 prompts in misinformation-prone areas (cancer, vaccines, stem cells, nutrition, athletic performance) were problematic. The chatbots were tested in February 2025. Source: BMJ Open.
• The February 2024 ransomware attack on Change Healthcare compromised the protected health information of approximately 192.7 million people, the largest healthcare data breach ever recorded. This was a ransomware attack, not an AI failure, included as a data security benchmark. Source: HHS Office for Civil Rights.
• A 2024 peer-reviewed study (Koenecke et al.) found OpenAI’s Whisper invented text in about 1.4% of transcriptions, and 38% of those contained explicit harms. Tests ran in April and May 2023 on aphasia research audio, so current versions may differ. Source: Koenecke et al., FAccT 2024..
• In a JAMA Health Forum analysis of 691 AI devices cleared through July 2023, 5.2% had reported adverse events (mostly malfunctions), including 1 death. These rates were lower than for non-AI 510(k) devices. Source: JAMA Health Forum.

Patient-Facing AI: Messaging and Portal Tools

• In a JAMA Network Open study, 16 primary care physicians assessed 344 pairs of AI- and human-written patient portal message replies without knowing which was which; accuracy, completeness, and tone did not differ statistically, and AI responses outperformed humans in understandability and tone by 9.5%. Source: Becker’s Hospital Review.
• In a 2024 Mass General Brigham study (The Lancet Digital Health) using GPT-4 and 100 simulated cancer patient questions, radiation oncologists judged AI-drafted replies safe in 82.1% of cases and acceptable to send without further editing in 58.3% of cases. If left unedited, 7.1% of responses could have posed a risk to the patient, and 0.6% could have posed a risk of death. Source: Mass General Brigham.
• Patients say they will accept AI-drafted portal messages only if a clinician reads every word before it reaches them, based on interviews with 40 patients from a large academic health system (JAMA Network Open, July 2026). Source: telehealth.org.

Physician Workforce and Capacity Data

• The AAMC (March 2024) projects a shortage of between 13,500 and 86,000 physicians in the US by 2036, driven largely by population aging. Source: AAMC.
• By 2034, Americans 65 and older are expected to outnumber children under 18 for the first time in US history, per AAMC workforce data.

RETURN ON INVESTMENT

• A Microsoft-commissioned IDC study published in March 2024 reported that 79% of healthcare organizations were using AI, with an average return of $3.20 per $1 invested and payback in 14 months. Microsoft sells healthcare AI, so treat this as a vendor-commissioned figure. (same link as above).
• For a deeper, independently verified look at real-world ROI numbers by use case, see our AI in Healthcare ROI: Independent Case Studies.

PATIENT PERSPECTIVE ON AI

• In a global survey of over 2,000 healthcare professionals and 20,000 patients across 10 countries, 62% of healthcare leaders said the benefits of AI investment are meeting or exceeding costs.
• 50% of healthcare professionals in the same survey said AI has increased their capacity to see more patients, and 58% of patients who used AI for health information said it helped them feel more prepared for appointments.(Philips Future Health Index 2026)

How We Verify These Numbers

Each figure above is checked against its original source (a peer-reviewed study, an official regulatory filing, or a primary organizational report) rather than a secondary aggregator, and linked directly so you can verify it yourself. Where a study has been revised, we cite the current version rather than an earlier draft. If you find an error or a newer figure, contact us and we’ll correct it.

FAQ

How many AI-enabled medical devices has the FDA approved?

The FDA had authorized 1,451 AI-enabled medical devices by the end of 2025, with 1,104 (76%) of them in radiology. Clearance activity is starting to spread into other specialties like cardiovascular, neurology, and pathology.

What percentage of doctors use AI in their practice?

81% of US physicians reported awareness or use of AI in their practice in 2026, more than double the 38% in 2023, according to the American Medical Association’s annual survey. The most common uses are summarizing medical research and drafting documentation, not autonomous diagnosis or treatment decisions.

How accurate is AI at diagnosing patients?

It depends heavily on the type of AI and the task. Narrow, purpose-built diagnostic tools perform very well: diabetic retinopathy screening reaches about 93% sensitivity, and one foundation-model CT triage tool reported 97% sensitivity in its company-reported pivotal study. General-purpose generative AI is far weaker at open-ended diagnosis, averaging only 52.1% accuracy across 83 studies, close to a non-expert clinician’s performance.

What are the biggest safety risks of AI in healthcare?

The two most documented risks are clinical safety and data security. In the NOHARM benchmark, general-purpose AI models carried potential for severe harm in up to 24.6% of cases, though specialized clinical tools performed dramatically better. Separately, the February 2024 Change Healthcare cyberattack (not an AI failure) compromised the health information of approximately 192.7 million people, the largest healthcare data breach ever recorded.

Related Reading

For deeper coverage of any category above, see our related pieces: 7 Real Risks of AI in Healthcare, AI in Healthcare Diagnosis, Will AI Replace Doctors?, and Generative AI in Healthcare: Tools and Adoption.