AI & data · sourced statistics
AI in Healthcare Statistics 2026
This page collects measured figures on artificial intelligence in medicine: how many hospitals and health systems use it, how many AI medical devices the FDA has authorized, how many physicians and clinicians use AI, what outcomes studies have found, how patients feel about it, and where the money is going. Most figures describe the United States, with UK figures from NHS England and the Health Foundation, European figures from the WHO Regional Office for Europe, and global clinician surveys from Elsevier and Philips. Data covers 2024 to 2026. Every figure carries the exact sentence it was taken from and a link to the page.
Compiled by the FruitPunch research desk · Updated November 15, 2026
- figures
- 53
- sources
- 20
- publishers
- 16
The short answer
AI is already common in U.S. hospitals: 71% used predictive AI built into their health record in 2024 (ASTP), and 22% of healthcare organizations had bought domain-specific AI tools by 2025 (Menlo Ventures). The FDA had authorized 1,357 AI medical devices by December 2025, 76.6% of them in radiology, with 258 in 2025 alone (Stanford AI Index 2026). Clinician use has nearly doubled, with 48% of clinicians using AI for work in 2025 (Elsevier), but 60% of U.S. adults remain uncomfortable with AI in their own diagnoses and treatments (Pew, 2026).
Quote these
Key findings
Each finding names who published it. Copy it with the source attached, or jump to it in context.
- 71%#1
In 2024, 71% of U.S. non-federal acute care hospitals reported using predictive AI integrated with their electronic health record, up from 66% in 2023.
Source: Assistant Secretary for Technology Policy (HHS), Sep 2025
In context - 22%#2
Menlo Ventures found that 22% of U.S. healthcare organizations had implemented domain-specific AI tools in 2025, a sevenfold increase over 2024.
Source: Menlo Ventures, Oct 2025
In context - 1,357#3
By December 2025 the FDA had authorized 1,357 AI and machine learning enabled medical devices from 693 companies across 17 clinical specialties.
Source: Stanford HAI, Apr 2026
In context - 76.6%#4
Radiology accounts for 76.6% of FDA-authorized AI devices, 1,039 of 1,357, followed by cardiovascular devices at 9.6%.
Source: Stanford HAI, Apr 2026
In context - 48%#5
In Elsevier's 2025 global survey, 48% of clinicians had used AI for work purposes, almost double the 26% reported a year earlier.
Source: Elsevier, Jul 2025
In context - 12%#6
In the Swedish MASAI randomized trial of more than 100,000 women, AI-supported mammography reduced the rate of breast cancer diagnosis in the years after screening by 12%.
Source: The Lancet, Jan 2026
In context - 51.9% to 38.8%#7
In a Yale-led study of 263 clinicians at six U.S. health systems, the share experiencing burnout fell from 51.9% to 38.8% after 30 days of using an AI scribe.
Source: Yale School of Medicine, Oct 2025
In context - 60%#8
In a June 2026 Pew Research Center survey, 60% of U.S. adults said they would be uncomfortable with AI being used for healthcare tasks like diagnoses and treatments, twice the 30% who were comfortable.
Source: Pew Research Center, Oct 2026
In context
Reading the numbers
Adoption depends on what you count. The federal health IT office (ASTP) reports that 71% of U.S. hospitals used predictive AI integrated with the electronic health record in 2024, but that includes models supplied by EHR vendors and switched on by default. Menlo Ventures measures something narrower, paid domain-specific AI tools, and finds 22% of healthcare organizations in 2025, led by health systems at 27%. Both show a wide divide: 86% of system-affiliated hospitals used predictive AI against 37% of independent ones.
FDA authorizations are rising fast. The Stanford AI Index 2026 counts 1,357 AI-enabled devices authorized by December 2025, with 258 in 2025 alone, and the Congressional Research Service cited about 1,450 by June 2026. Radiology accounts for 76.6% of devices. The evidence base is thin, since a peer-reviewed analysis found only 2.4% of devices with clinical studies were supported by randomized trial data.
Clinicians are moving quickly, especially toward documentation tools. Elsevier finds 48% of clinicians used AI for work in 2025, up from 26%, and Stanford names ambient scribes the most widely adopted clinical AI category. NHS England reports 23.5% more direct patient time with AI scribes, Kaiser Permanente counts the equivalent of 1,794 working days saved in one year, and a Yale-led study saw burnout fall from 51.9% to 38.8%. The clearest clinical result comes from screening: the MASAI randomized trial in The Lancet found AI-supported mammography cut later interval cancers by 12%.
The public is more cautious than clinicians. Pew finds 60% of U.S. adults uncomfortable with AI in their own diagnoses and treatments, and 47% expect it to harm the patient and doctor relationship. At the same time, people use chatbots for health on their own: KFF finds 29% sought physical health information from AI in the past year, and OpenAI reports more than 40 million daily health questions on ChatGPT. Governance lags, with only 4 of 50 European countries holding a health-specific AI strategy, according to WHO Europe.
How many hospitals and health systems use AI
- 171%
In 2024, 71% of U.S. non-federal acute care hospitals reported using predictive AI integrated with their electronic health record, up from 66% in 2023.
Source: Assistant Secretary for Technology Policy (HHS), Sep 2025
Exact wording at the source
In 2024, 71% of hospitals reported using predictive AI integrated with the electronic health record (EHR), up from 66% in 2023.
- 286% vs. 37%
System-affiliated U.S. hospitals used predictive AI far more often than independent hospitals in 2024: 86% compared with 37%.
Source: Assistant Secretary for Technology Policy (HHS), Sep 2025
Exact wording at the source
For example, system-affiliated hospitals used predictive AI at significantly higher rates than independent hospitals in both 2023 (81% vs. 31%) and 2024 (86% vs. 37%).
- 390%
Among U.S. hospitals using the market-leading EHR vendor, 90% used predictive AI in 2024, compared with 50% of hospitals on other vendors.
Source: Assistant Secretary for Technology Policy (HHS), Sep 2025
Exact wording at the source
Predictive AI use also varied by EHR vendor: 90% of hospitals using the market leading EHR vendor used predictive AI in 2024 compared to just 50% of hospitals using the other vendors.
- 482%
Of U.S. hospitals using predictive AI in 2024, 82% evaluated models for accuracy, 74% for bias and 79% monitored them after implementation.
Source: Assistant Secretary for Technology Policy (HHS), Sep 2025
Exact wording at the source
In 2024, 82% of hospitals evaluated predictive AI for accuracy, 74% for bias, and 79% conducted post-implementation evaluation or monitoring.
- 522%
Menlo Ventures found that 22% of U.S. healthcare organizations had implemented domain-specific AI tools in 2025, a sevenfold increase over 2024.
Source: Menlo Ventures, Oct 2025
Exact wording at the source
According to Menlo Ventures’ research, 22% of healthcare organizations have implemented domain-specific AI tools, a 7x increase over 2024 and 10x over 2023.
- 627%
Health systems had the highest adoption of domain-specific AI tools in Menlo Ventures' 2025 survey at 27%, ahead of outpatient providers and payers.
Source: Menlo Ventures, Oct 2025
Exact wording at the source
Health systems are leading the charge at 27%, outpacing outpatient facilities (18%) and payers (14%).
- 72.2x
According to Menlo Ventures, U.S. healthcare organizations deployed AI at 2.2 times the rate of the broader U.S. economy in 2025.
Source: Menlo Ventures, Oct 2025
Exact wording at the source
Healthcare organizations are deploying AI at 2.2x the rate of the broader U.S.
- 86.6 months
U.S. health systems shortened their average buying cycle for AI to 6.6 months in 2025, compared with 8.0 months for traditional IT purchases.
Source: Menlo Ventures, Oct 2025
Exact wording at the source
Health systems have shortened average buying cycles from 8.0 months for traditional IT purchases to 6.6 months, an 18% acceleration.
- 9Broadest adoption
Ambient AI scribes saw the broadest adoption of any clinical AI category in 2025, according to the Stanford AI Index 2026.
Source: Stanford HAI, Apr 2026
Exact wording at the source
Ambient AI scribes, tools that automatically generate clinical documentation from patient–clinician conversations, saw the broadest adoption of any clinical AI category in 2025.
Chart
Healthcare organizations with domain-specific AI tools by segment, 2025
Menlo Ventures and Morning Consult survey of U.S. healthcare executives, 2025.
- Health systems27%
- Outpatient providers18%
- Payers14%
Source: Menlo Ventures, 2025: The State of AI in Healthcare
FDA-authorized AI medical devices
- 101,357
By December 2025 the FDA had authorized 1,357 AI and machine learning enabled medical devices from 693 companies across 17 clinical specialties.
Source: Stanford HAI, Apr 2026
Exact wording at the source
By December 2025, the FDA had authorized a total of 1,357 AI/ML-enabled medical devices from 693 different companies across 17 clinical specialties (Figure 6.2.6).
- 11258
The FDA authorized 258 AI medical devices in 2025 through September alone, already more than in any earlier full year.
Source: Stanford HAI, Apr 2026
Exact wording at the source
Annual authorizations reached 258 through September 2025, already surpassing all prior full-year totals.
- 1276.6%
Radiology accounts for 76.6% of FDA-authorized AI devices, 1,039 of 1,357, followed by cardiovascular devices at 9.6%.
Source: Stanford HAI, Apr 2026
Exact wording at the source
Radiology accounts for the largest share of authorized AI/ML devices at 1,039 of 1,357 (76.6%), followed by cardiovascular (130 devices, 9.6%) and neurology (61 devices, 4.5%) (Figure 6.2.7).
- 13246
The number of AI and machine learning devices cleared through the FDA 510(k) pathway reached 246 in 2025, up from 16 in 2016.
Source: Stanford HAI, Apr 2026
Exact wording at the source
The number of 510(k)-cleared AI/ML-related devices reached 246 in 2025, continuing a steep upward trajectory that began with 16 devices in 2016 (Figure 6.2.5).
- 142.4%
In a peer-reviewed analysis of all 1,016 FDA AI device authorizations through December 2024, only 2.4% of devices with clinical studies were supported by randomized controlled trial data.
Source: Stanford HAI, Apr 2026
Exact wording at the source
Despite this growth, a peer-reviewed analysis of all 1,016 authorizations through December 2024 (Singh et al., 2025) found that only 2.4% of devices with clinical studies were supported by randomized controlled trial data, with nearly all devices entering via the 510(k) pathway.
- 151,450
As of June 2026, the Congressional Research Service reported that about 1,450 AI-enabled devices had been authorized for marketing by the FDA, mostly in radiology, cardiology and neurology.
Source: Congressional Research Service, Jun 2026
Exact wording at the source
FDA reports that approximately 1,450 such devices have been authorized for marketing, with the majority in the fields of radiology, cardiology, and neurology.
How many physicians and clinicians use AI
- 1648%
In Elsevier's 2025 global survey, 48% of clinicians had used AI for work purposes, almost double the 26% reported a year earlier.
Source: Elsevier, Jul 2025
Exact wording at the source
In 2025, 76% of clinicians have used an AI tool, and nearly half (48%) have used AI specifically for work purposes, almost double the 26% reported just a year earlier.
- 1797%
Among clinicians who used AI in Elsevier's 2025 survey, 97% used generalist tools such as ChatGPT, and 76% of work users had tried clinical-specific tools.
Source: Elsevier, Jul 2025
Exact wording at the source
While generalist AI tools like ChatGPT are the most widely used (97% of AI users), clinical-specific AI tools are also gaining traction, with 76% of work-related AI users having tried them.
- 1857%
In Elsevier's 2025 survey, 57% of clinicians said clinical AI tools save them time.
Source: Elsevier, Jul 2025
Exact wording at the source
Over half of clinicians (57%) say that clinical AI tools save them time, and 53% feel empowered by these technologies.
- 1930%
Only 30% of clinicians surveyed by Elsevier in 2025 said they receive enough AI training from their institution.
Source: Elsevier, Jul 2025
Exact wording at the source
Only a third (32%) feel supported with digital tools like AI from their institution, and few receive enough AI training from their institution (30%).
- 2040%
OpenEvidence, an AI evidence retrieval platform, reported adoption by 40% of U.S. physicians, according to the Stanford AI Index 2026.
Source: Stanford HAI, Apr 2026
Exact wording at the source
OpenEvidence, a real-time evidence retrieval platform, reported adoption by 40% of U.S. physicians.
- 2163%
In Philips' Future Health Index 2025, 63% of healthcare professionals across 16 countries were optimistic that AI could improve patient outcomes.
Source: Philips, May 2025
Exact wording at the source
While a majority of HCPs (63%) are optimistic that AI could improve patient outcomes, less than half of patients surveyed are optimistic that AI can improve healthcare.
- 2285%
In the same Philips survey, 85% of healthcare professionals said AI can reduce their administrative burden.
Source: Philips, May 2025
Exact wording at the source
Moreover, 85% state that AI can reduce their administrative burden, and 74% see AI improving patient access through increased capacity and throughput as well as increased face-to-face time with patients, reinforcing the crucial role AI can play in speeding healthcare delivery.
- 2376%
In a 2024 Health Foundation survey, 76% of NHS staff supported the use of AI for patient care and 81% for administrative work.
Source: The Health Foundation, Jul 2024
Exact wording at the source
Support for the use of AI was even greater among the NHS staff surveyed, with 76% supporting its use for patient care and 81% for administrative purposes.
- 247,260
At Kaiser Permanente in Northern California, 7,260 physicians used AI scribes in more than 2.5 million patient encounters between October 2023 and December 2024.
Source: Kaiser Permanente, Jun 2025
Exact wording at the source
AI scribes were used by 7,260 Permanente physicians in more than 2.5 million patient encounters during the evaluation period.
Chart
Clinicians using AI for work, 2024 and 2025
Elsevier Clinician of the Future global surveys of doctors and nurses.
- 202426%
- 202548%
Source: Elsevier, How AI is revolutionizing clinician workflows: insights from 2,000 global clinicians
Clinical outcomes and time savings
- 2512%
In the Swedish MASAI randomized trial of more than 100,000 women, AI-supported mammography reduced the rate of breast cancer diagnosis in the years after screening by 12%.
Source: The Lancet, Jan 2026
Exact wording at the source
Artificial intelligence (AI)-supported mammography identifies more cancers during screening and reduces the rate of breast cancer diagnosis by 12% in the years following, finds the first randomised controlled trial of its kind involving over 100,000 Swedish women published in The Lancet journal.
- 2644%
Interim results of the MASAI trial found a 44% reduction in screen-reading workload for radiologists.
Source: The Lancet, Jan 2026
Exact wording at the source
The interim safety results of the MASAI trial, published in The Lancet Oncology in 2023 [1], found a 44% reduction in screen-reading workload for radiologists.
- 2723.5%
In an NHS evaluation of more than 17,000 patient encounters, AI scribes increased direct patient interaction time by 23.5% and cut appointment length by 8.2%.
Source: NHS England, Jan 2026
Exact wording at the source
Results showed a 23.5% increase in direct patient interaction time during appointments, alongside an 8.2% reduction in overall appointment length when AI-scribes were used.
- 2813.4%
Emergency departments in the same NHS evaluation saw 13.4% more patients per shift when AI scribes were used.
Source: NHS England, Jan 2026
Exact wording at the source
A&E saw particularly strong results, with a 13.4% increase in patients seen per shift.
- 2951.9% to 38.8%
In a Yale-led study of 263 clinicians at six U.S. health systems, the share experiencing burnout fell from 51.9% to 38.8% after 30 days of using an AI scribe.
Source: Yale School of Medicine, Oct 2025
Exact wording at the source
Results showed that the proportion of clinicians experiencing burnout dropped from 51.9% to 38.8% , with participants reporting less time spent documenting after hours, lower cognitive load, and greater ability to focus attention on patients.
- 301,794 working days
AI scribes saved Kaiser Permanente physicians in Northern California the equivalent of 1,794 working days in one year.
Source: Kaiser Permanente, Jun 2025
Exact wording at the source
A new analysis published in NEJM Catalyst shows that this innovation saved Permanente physicians in Northern California the equivalent of 1,794 working days in one year — nearly five years of work hours.
- 3184%
Kaiser Permanente physicians said AI scribes had a positive effect on patient interactions (84%) and overall work satisfaction (82%).
Source: Kaiser Permanente, Jun 2025
Exact wording at the source
Physicians overwhelmingly said the technology had a positive effect on patient interactions (84%) and overall work satisfaction (82%).
- 3283%
Sharp HealthCare reported an 83% reduction in note-writing effort after deploying an ambient AI documentation tool.
Source: Stanford HAI, Apr 2026
Exact wording at the source
Sharp HealthCare reported an 83% reduction in note-writing effort and a 3.5%–6% increase in work relative value units—a standard measure of physician clinical productivity—per encounter.
- 3318.7%
An AI sepsis early warning system deployed across 13 Cleveland Clinic hospitals was linked to an 18.7% relative reduction in sepsis mortality.
Source: Stanford HAI, Apr 2026
Exact wording at the source
Reported outcomes included an 18.7% relative reduction in sepsis mortality, a 1.85-hour reduction in median time to first antibiotic order, the correct identification of 82% of sepsis cases, an 89% clinician adoption rate, and a 10% reduction in intensive care unit utilization.
- 3485.5%
Microsoft's MAI-DxO multi-agent system scored 85.5% on complex published case studies, versus 20% for unaided physicians.
Source: Stanford HAI, Apr 2026
Exact wording at the source
A multi-agent AI system scored 85.5% on complex published case studies, versus 20% for unaided physicians.
Patients, the public and AI for health
- 3560%
In a June 2026 Pew Research Center survey, 60% of U.S. adults said they would be uncomfortable with AI being used for healthcare tasks like diagnoses and treatments, twice the 30% who were comfortable.
Source: Pew Research Center, Oct 2026
Exact wording at the source
A majority (60%) is uncomfortable with AI being used for healthcare tasks like diagnoses and treatments, double the share who says they’re comfortable (30%).
- 3637%
In 2026, 37% of U.S. adults told Pew that AI for diagnoses and treatments will lead to better health outcomes, while 36% expected worse outcomes.
Source: Pew Research Center, Oct 2026
Exact wording at the source
Some 37% of Americans say the use of AI for things like diagnoses and treatments will lead to better health outcomes.
- 3747%
Pew found that 47% of U.S. adults in 2026 expect AI to make the relationship between patients and doctors worse, compared with 10% who expect it to improve.
Source: Pew Research Center, Oct 2026
Exact wording at the source
And at a time when many see AI eroding human connection in general, 47% say AI will make the relationship between patients and their doctors worse, compared with 10% who say it will make it better.
- 3822%
In 2026, 22% of U.S. adults told Pew they get health information from AI chatbots at least sometimes, compared with 36% from social media.
Source: Pew Research Center, Apr 2026
Exact wording at the source
Still, 36% of Americans say they at least sometimes get health information from social media, and 22% say the same about AI chatbots.
- 3929%
In a KFF poll from early 2026, 29% of U.S. adults had used AI for information about their physical health and 16% about their mental health in the past year.
Source: KFF, Mar 2026
Exact wording at the source
The share using AI for health advice includes about 3 in 10 (29%) who have sought information related to their physical health and about 1 in 6 (16%) who have sought information related to their mental health.
- 4041%
Among U.S. adults who use AI for health information, 41% have uploaded personal medical information such as test results or doctors' notes, according to KFF.
Source: KFF, Mar 2026
Exact wording at the source
Among those who use AI for health information, 41% say that they have uploaded personal medical information like test results or doctors’ notes into an AI tool or chatbot to get personalized explanations or advice related to their health.
- 4177%
KFF found that 77% of U.S. adults are concerned about the privacy of personal medical information given to AI tools.
Source: KFF, Mar 2026
Exact wording at the source
Among the public at large, about three-quarters (77%) say that they are concerned about the privacy of personal medical information provided to AI tools, including most (65%) of those who have shared personal medical information with AI.
- 42One in four
OpenAI reports that one in four of its more than 800 million regular ChatGPT users submits a prompt about healthcare every week.
Source: OpenAI, Jan 2026
Exact wording at the source
Of our more than 800 million regular users, one in four submits a prompt about healthcare every week.
- 4340 million
More than 40 million people turn to ChatGPT every day with healthcare questions, according to OpenAI.
Source: OpenAI, Jan 2026
Exact wording at the source
More than 40 million turn to ChatGPT every day with healthcare questions.
- 4454%
In a 2024 Health Foundation survey, 54% of the UK public supported the use of AI for patient care such as diagnosing illness and recommending treatment.
Source: The Health Foundation, Jul 2024
Exact wording at the source
Over half of the UK public (54%) and three-quarters of NHS staff surveyed (76%) said they support the use of AI for patient care, and an even greater proportion said they support the use of AI for administrative purposes (61% of the public and 81% of NHS staff surveyed).
Chart
U.S. adults' comfort with their doctor using AI for diagnoses and treatments, 2026
Pew Research Center survey of 3,488 U.S. adults, June 2026.
- Uncomfortable60%
- Comfortable30%
Source: Pew Research Center, Americans Are Open to AI in Healthcare, but Want Humans Making Final Calls
Spending, investment and national readiness
- 45$1.4 billion
Menlo Ventures estimates U.S. healthcare AI spending reached $1.4 billion in 2025, nearly triple the 2024 level.
Source: Menlo Ventures, Oct 2025
Exact wording at the source
Healthcare AI spending hit $1.4 billion this year
- 4685%
According to Menlo Ventures, 85% of generative AI spending in U.S. healthcare in 2025 went to startups rather than incumbents.
Source: Menlo Ventures, Oct 2025
Exact wording at the source
Our survey reveals that 85% of all generative AI spend in healthcare currently flows to startups rather than incumbents.
- 47$600 million
Ambient scribes generated $600 million in U.S. healthcare AI revenue in 2025, the largest clinical AI category, according to Menlo Ventures.
Source: Menlo Ventures, Oct 2025
Exact wording at the source
Ambient scribes are healthcare AI’s first breakout category, generating $600 million in 2025
- 4854%
AI-enabled digital health companies captured 54% of U.S. digital health venture funding in 2025, up from 37% in 2024, according to Rock Health.
Source: Rock Health, Jan 2026
Exact wording at the source
AI-enabled digital health companies captured 54% of total funding (up from 37% last year) and commanded a roughly 19% premium on average deal size compared to companies not explicitly centering AI in their products and services.
- 4950%
Half of U.S. digital health venture deals in 2025 were closed by AI-enabled companies, Rock Health found.
Source: Rock Health, Jan 2026
Exact wording at the source
Across 2025, 50% of deals were closed by “AI-enabled” companies, 3 matching the broader VC market.
- 50£10 billion
NHS England is allocating £10 billion over three years to technology, digital and data, including AI triage in the NHS App and AI notetaking.
Source: NHS England, Jul 2026
Exact wording at the source
Today the NHS is setting out how £10 billion of funding over the next three years, allocated by the government last year, will be used for a major overhaul of the health service’s technology, digital and data systems.
- 5164%
In a WHO Europe survey of 50 countries, 64% already use AI-assisted diagnostics, particularly in imaging.
Source: WHO Regional Office for Europe, Nov 2025
Exact wording at the source
And yet, there are reasons for optimism: 64% (32 out of 50) of countries already use AI-assisted diagnostics, particularly in imaging.
- 524 of 50
Only 4 of 50 countries surveyed by WHO Europe have a health-specific national AI strategy.
Source: WHO Regional Office for Europe, Nov 2025
Exact wording at the source
The findings are sobering: Only 4 out of 50 countries have a health-specific national AI strategy.
- 5386%
In the WHO Europe survey, 86% of countries cited legal uncertainty as their top barrier to AI adoption in health, followed by financial constraints at 78%.
Source: WHO Regional Office for Europe, Nov 2025
Exact wording at the source
86% (or 43 out of 50) Member States cite legal uncertainty as their top barrier to adoption, followed closely by financial constraints (78%; 39 out of 50).
Chart
Share of U.S. digital health venture funding going to AI-enabled companies
Rock Health digital health funding database, U.S. venture deals of $2 million or more.
- 202437%
- 202554%
Source: Rock Health, 2025 year-end digital health funding overview: A tale of two markets
FAQ
Questions readers ask
- How many hospitals use AI?
- About 7 in 10 U.S. hospitals. The Assistant Secretary for Technology Policy reports that 71% of non-federal acute care hospitals used predictive AI integrated with their electronic health record in 2024, up from 66% in 2023. Paid domain-specific AI tools are less common: Menlo Ventures counts 22% of healthcare organizations in 2025.
- How many AI medical devices has the FDA approved?
- The Stanford AI Index 2026 counts 1,357 FDA-authorized AI and machine learning devices by December 2025, including 258 in 2025. The Congressional Research Service cited about 1,450 as of June 2026. Radiology makes up 76.6% of them.
- What percentage of doctors use AI?
- Elsevier's 2025 global survey found 48% of clinicians used AI for work, almost double the 26% a year earlier. Stanford's AI Index reports that one evidence retrieval tool, OpenEvidence, is used by 40% of U.S. physicians.
- Does AI improve patient outcomes?
- The best evidence is in screening and documentation. The MASAI randomized trial in The Lancet found AI-supported mammography reduced later breast cancer diagnoses by 12%, and a Yale-led study saw clinician burnout fall from 51.9% to 38.8% with AI scribes. Most other results are observational.
- Are patients comfortable with AI in healthcare?
- Mostly not for clinical decisions. Pew found in 2026 that 60% of U.S. adults would be uncomfortable with AI being used for diagnoses and treatments, and KFF found 77% are concerned about the privacy of medical information shared with AI tools.
Sources
We chose figures from government bodies (Assistant Secretary for Technology Policy at HHS, Congressional Research Service, NHS England, WHO Regional Office for Europe), research institutes and journals (Stanford HAI, The Lancet, Yale School of Medicine, the Health Foundation), survey organizations (Pew Research Center, KFF) and the companies that ran each survey or analyzed their own data (Menlo Ventures, Rock Health, Elsevier, Philips, Kaiser Permanente, OpenAI). Definitions differ: hospital predictive AI counts models embedded in the EHR, while Menlo counts paid domain-specific tools; FDA device totals vary by list version and cutoff date. Several scribe studies are observational, and some vendor data is self-reported. The American Medical Association's 2026 physician survey could not be fetched for verification, so its figures are not included. We store the sentence each number appears in, and a script reloads the publisher's page to confirm it is still there. The methodology has the details.
- Assistant Secretary for Technology Policy (HHS). Hospital Trends in the Use, Evaluation, and Governance of Predictive AI, 2023-2024 (Sep 2025).
- Congressional Research Service. FDA Regulation of AI-Enabled Devices (Jun 2026).
- Elsevier. How AI is revolutionizing clinician workflows: insights from 2,000 global clinicians (Jul 2025).
- Elsevier. Clinician of the Future 2025 (Jul 2025).
- Kaiser Permanente. Analysis: AI scribes save physicians time, improve patient interactions and work satisfaction (Jun 2025).
- KFF. Poll: 1 in 3 Adults Are Turning to AI Chatbots for Health Information, Equaling the Share Who Use Social Media for Health (Mar 2026).
- Menlo Ventures. 2025: The State of AI in Healthcare (Oct 2025).
- NHS England. NHS backs AI notetaking to free up more face-to-face care (Jan 2026).
- NHS England. NHS accelerates artificial intelligence rollout to cut waiting times and improve care for millions (Jul 2026).
- OpenAI. AI as a Healthcare Ally: How Americans are navigating the system with ChatGPT (Jan 2026).
- Pew Research Center. Americans Are Open to AI in Healthcare, but Want Humans Making Final Calls (Oct 2026).
- Pew Research Center. Where Do Americans Get Health Information, and What Do They Trust? (Apr 2026).
- Philips. Philips Future Health Index 2025 report highlights significant trust gap in AI (May 2025).
- Rock Health. 2025 year-end digital health funding overview: A tale of two markets (Jan 2026).
- Stanford HAI. AI Index Report 2026, Chapter 6: Medicine (Apr 2026).
- Stanford HAI. Medicine | The 2026 AI Index Report (Apr 2026).
- The Health Foundation. AI in health care: what do the public and NHS staff think? (Jul 2024).
- The Lancet. AI-supported mammography screening results in fewer aggressive and advanced breast cancers, finds full results from first randomized controlled trial (Jan 2026).
- WHO Regional Office for Europe. Statement: Humanity must hold the pen: the European Region can write the story of ethical AI for health (Nov 2025).
- Yale School of Medicine. Study: AI Scribes Lower Physician Burnout, Lend Patients More Undivided Attention (Oct 2025).
Cite this page
Link to this page or to a single numbered figure. If you reuse a number, name the organization that first published it.
FruitPunch. (2026, November 15). AI in Healthcare Statistics 2026. https://fruitpunch.ai/ai-in-healthcare-statistics
Found a number that changed or a source that moved? Email editorial@fruitpunch.ai. We plan to check this page again by January 14, 2027.
Related rankings
- Best AI Chatbots · top pick ChatGPT
Keep reading