Skip to content
ReviewRamble

The weekly brief

AI, reviewed, in your inbox.

  • New reviews of tools, models, MCP servers and repos, with the verdict up front
  • Re-checked prices and ratings when something changes
  • New analysis pieces as they are published

Models, agents, MCP servers and the tools actually worth your time.

One email a week. Unsubscribe with the link in any brief. How we use your email

Research

AI Adoption Statistics 2026

How many businesses, workers, consumers and developers actually use AI? This page collects 123 figures from 22 primary sources, each linked to the original and labelled with its own measurement period.

By ReviewRamble editors · Reviewed Nov 18, 2026 · Methodology and definitions

Key takeaways

  • About one in five businesses uses AI by official measures: 19.8% in the U.S. (May 2026) and 19.95% in the EU (2025). Surveys of large organizations report 88%, so the difference is who gets counted, not a contradiction.
  • Workers are ahead of their employers: 56% of U.S. workers have used AI on the job, and 55% of UK employees use it for work or education, but daily use is still a minority habit.
  • Developers use AI the most and trust it the least: 90% use AI tools at work, while 46% distrust the accuracy of the output.

US business adoption

The U.S. Census Bureau's Business Trends and Outlook Survey (BTOS) asks a nationally representative sample of employer businesses every two weeks whether they used AI in the previous two weeks. It is the broadest measure of U.S. business adoption, because it counts small firms as well as large ones.

In November 2025 Census changed the question from AI used "in producing goods or services" to AI used "in any business function", so rates before and after that date are not directly comparable.

US business AI use by firm size and sector

Share of U.S. employer businesses using AI, period ending May 3, 2026

All businesses19.8%250+ employees37%100-249 employees32%Information39.7%Finance and insurance33.9%Retail trade14%
BTOS; "any business function" wording from Nov 2025; 4-or-fewer is an upper bound (<20%). Chart by ReviewRamble. Data: U.S. Census Bureau.
  1. 19.8% of U.S. businesses used AI in a business function in the two weeks ending May 3, 2026. U.S. Census Bureau
  2. Between 17% and 20% of U.S. businesses reported using AI in each BTOS wave from December 2025 to May 2026. U.S. Census Bureau
  3. Between 20% and 23% of U.S. businesses expected to be using AI within the next six months over the same December 2025 to May 2026 waves. U.S. Census Bureau
  4. 37% of U.S. firms with at least 250 employees reported using AI in their business operations (BTOS, December 2025 to May 2026 review). U.S. Census Bureau
  5. 32% of U.S. firms with 100 to 249 employees said they used AI in the collection period ending May 3, 2026. U.S. Census Bureau
  6. Fewer than 20% of U.S. firms with four or fewer employees reported using AI, and adoption among firms under 20 employees did not change significantly between December 2025 and May 2026. U.S. Census Bureau
  7. 39.7% of businesses in the Information sector used AI as of May 3, 2026, about twice the national rate. U.S. Census Bureau
  8. 33.9% of Finance and Insurance businesses used AI as of May 3, 2026. U.S. Census Bureau
  9. About 42% of Information-sector businesses expected to use AI in a business function over the next six months (May 2026). U.S. Census Bureau
  10. Around 14% of Retail Trade businesses used AI, below the national average (May 2026). U.S. Census Bureau
  11. 18% of U.S. firms used AI in a business function during the BTOS AI supplement reference period, November 2025 to January 2026. U.S. Census Bureau, Center for Economic Studies
  12. 32% of U.S. firms used AI when weighted by employment, meaning about a third of jobs are in AI-using firms (Nov 2025 to Jan 2026). U.S. Census Bureau, Center for Economic Studies
  13. 57% of AI-using U.S. firms integrate AI in three or fewer business functions (Nov 2025 to Jan 2026). U.S. Census Bureau, Center for Economic Studies
  14. Sales and Marketing is the most common business function for AI, at 52% of AI-using firms, ahead of Strategy and Business Development (45%) and IT (41%). U.S. Census Bureau, Center for Economic Studies
  15. 66% of AI-using U.S. firms rely on AI only to augment tasks, not to replace them (Nov 2025 to Jan 2026). U.S. Census Bureau, Center for Economic Studies
  16. Only 2% of U.S. firms reported AI-related employment decreases, concentrated among large firms (Nov 2025 to Jan 2026). U.S. Census Bureau, Center for Economic Studies

What this shows: The size gradient is the story: firms with 250 or more employees report roughly twice the national rate, and adoption among the smallest firms barely moved over the six months. Employment-weighted, about a third of U.S. jobs sit in AI-using firms even though fewer than one in five firms use AI.

Business adoption in the EU and UK

Eurostat's annual ICT usage survey counts enterprises with 10 or more employees that used at least one of a list of AI technologies during the year. The UK Office for National Statistics asks a similar question every two weeks in its Business Insights and Conditions Survey (BICS).

Both exclude the smallest firms, which the U.S. BTOS includes, and BICS also leaves out finance and insurance, agriculture and the public sector. Compare these figures with each other, not with the U.S. rate.

EU enterprises using AI, 2025

Share of enterprises with 10+ employees using at least one AI technology

EU average19.95%Denmark42.03%Finland37.82%Sweden35.04%Poland8.36%Romania5.21%Small (10-49)17%Medium (50-249)30.36%Large (250+)55.03%
Eurostat ICT usage in enterprises survey, 2025; selected countries and size classes. Chart by ReviewRamble. Data: Eurostat.
  1. 19.95% of EU enterprises with 10 or more employees used at least one AI technology in 2025. Eurostat
  2. EU enterprise AI use rose 6.47 percentage points between 2024 and 2025. Eurostat
  3. 55.03% of large EU enterprises (250+ employees) used AI in 2025, against 30.36% of medium and 17% of small enterprises. Eurostat
  4. Denmark had the highest share of enterprises using AI in the EU, 42.03%, followed by Finland (37.82%) and Sweden (35.04%) in 2025. Eurostat
  5. Romania had the lowest share, 5.21% of enterprises, followed by Poland (8.36%) and Bulgaria (8.55%) in 2025. Eurostat
  6. 62.52% of EU enterprises in information and communication used AI in 2025, the highest of any economic activity. Eurostat
  7. 40.43% of EU enterprises in professional, scientific and technical activities used AI in 2025; every other activity was below 25%. Eurostat
  8. 34.70% of AI-using EU enterprises used it for marketing or sales, the most common purpose in 2025. Eurostat
  9. 14.21% of EU enterprises that did not use AI had considered using it in 2025. Eurostat
  10. 70.89% of EU enterprises that considered AI but did not use it cited lack of relevant expertise, ahead of unclear legal consequences (52.52%). Eurostat
  11. About 35% of UK businesses with 10 or more employees used AI by June 2026, up from about 12% in late 2023. Office for National Statistics
  12. Large language models were the most used AI technology, at 18% of UK businesses with 10+ employees, followed by visual content creation (16%) in June 2026. Office for National Statistics
  13. 58% of UK information and communication businesses used AI, against 13% in construction (June 2026). Office for National Statistics
  14. Only 10% of UK businesses using AI said they use it extensively (10+ employees, June 2026). Office for National Statistics
  15. Only 11% of UK businesses said more than half their workforce had received AI-related training (10+ employees, June 2026). Office for National Statistics
  16. 41% of UK businesses with 10+ employees reported no barriers to adopting AI in the last three months (June 2026). Office for National Statistics
  17. 26% of all UK businesses used at least one AI technology in late March 2026, up 8 percentage points in a year. Office for National Statistics
  18. 45% of UK businesses with 250 or more employees used AI in late March 2026, up 13 percentage points in a year. Office for National Statistics
  19. 18% of UK businesses planned to adopt an AI technology in the next 3 months, the highest since the question began in September 2023 (late March 2026). Office for National Statistics
  20. 5% of UK businesses using AI reported a reduction in headcount because of AI in late March 2026 (7% among firms with 10+ employees). Office for National Statistics

What this shows: Country spread matters more than the EU average: Denmark is eight times Romania. In both the EU and the UK, the stated barrier is skills rather than money, and only a small share of users describe their AI use as extensive.

Why AI adoption estimates range from 18% to 88%

Writers often set an 88% adoption figure next to a 19.8% one as if they measured the same thing. They do not. The gap comes from who is surveyed (all firms, firms with 10+ employees, or mostly large organizations), whose answer counts (a firm, an employee, or a person) and how a figure is weighted (each firm equal, or weighted by employment).

Same question, different populations

Headline "AI adoption" figures by who was asked

US firms (BTOS)19.8%EU firms 10+ (Eurostat)19.95%US firms, emp-weighted32%UK firms 10+ (ONS)35%US employees: org adopted47%UK employees using AI55%Organizations (AI Index)88%
Each bar counts a different population, so this is not a like-for-like comparison. Chart by ReviewRamble. Data: U.S. Census Bureau, U.S. Census Bureau, Center for Economic Studies, Eurostat, Office for National Statistics, Gallup, Stanford HAI.
  1. Organizational AI adoption reached 88% in the survey the 2026 AI Index reports, which covers organizations rather than all businesses (2025). Stanford HAI
  2. Generative AI reached 53% population adoption within three years of launch, faster than the PC or the internet, according to the 2026 AI Index. Stanford HAI
  3. The U.S. ranks 24th for generative AI adoption, at 28.3% of the population, behind Singapore (61%) and the United Arab Emirates (64%), per the 2026 AI Index. Stanford HAI
  4. 55% of UK employees say they use AI for work or education, compared with about 35% of UK businesses (10+ employees) that report using at least one AI technology. Office for National Statistics
  5. 47% of U.S. employees say their organization has integrated AI tools to improve productivity, efficiency or quality, up from 41% the previous quarter (Q2 2026). Gallup
  6. In 23% of U.S. firms, workers use AI in work tasks (41% employment-weighted), a different measure from firm-level adoption (Nov 2025 to Jan 2026). U.S. Census Bureau, Center for Economic Studies
  7. U.S. firm adoption was expected to reach 22% within six months of the November 2025 to January 2026 supplement period. U.S. Census Bureau, Center for Economic Studies
  8. Use rates reach 50% to 60% for very large firms in the Information, Professional Services and Finance sectors (60% to 70% employment-weighted), Nov 2025 to Jan 2026. U.S. Census Bureau, Center for Economic Studies

What this shows: Rule of thumb for writers: quote the official business rate (BTOS 19.8%, Eurostat 19.95%, ONS 26% all firms or about 35% at 10+ employees) when you mean "businesses", an employee survey when you mean "workers", and label large-organization surveys as such. Never set the 88% next to an official rate without saying the populations differ.

Workers and AI at work

Two kinds of source measure AI at work: government household surveys (Census's Household Trends and Outlook Pulse Survey, HTOPS) and Gallup's quarterly panel of U.S. employees. Both depend on what people report about themselves. Time saved is the worker's own estimate, not measured productivity.

Time AI saved U.S. workers last week

Share of workers who used AI at work in the last week, by hours saved

No time saved10%Less than 1 hour25%1-2 hours31%3-4 hours15%More than 4 hours15%
Census HTOPS, March 2026; self-reported; 4% said AI added time (not shown). Chart by ReviewRamble. Data: U.S. Census Bureau.
  1. About 56% of U.S. workers have used AI on the job for at least one of 11 tasks (March 2026). U.S. Census Bureau
  2. 37% of U.S. workers have used AI to search for information or technical help, the most common of the 11 tasks (March 2026). U.S. Census Bureau
  3. 32% of U.S. workers have used AI to write communications, documentation or instructions (March 2026). U.S. Census Bureau
  4. 27% of U.S. workers have used AI for administrative tasks (March 2026). U.S. Census Bureau
  5. About 24% of U.S. workers who use AI at work used it every day in the last week (March 2026). U.S. Census Bureau
  6. 31% of recent AI users said AI saved them one to two hours of work in the last week; a quarter said less than an hour (March 2026). U.S. Census Bureau
  7. 15% of recent AI users said AI saved three to four hours last week, and another 15% said more than four hours (March 2026). U.S. Census Bureau
  8. 10% of recent AI users said AI saved no time, and 4% said it made the work take longer (March 2026). U.S. Census Bureau
  9. 30% of male workers who use AI used it every day last week, against 17% of female workers (March 2026). U.S. Census Bureau
  10. 53% of U.S. employees use AI in their role at least a few times a year (Q3 2026). Gallup
  11. 32% of U.S. employees use AI frequently (a few times a week or more), up two percentage points from the previous quarter (Q3 2026). Gallup
  12. 16% of U.S. employees use AI daily (Q3 2026). Gallup
  13. 49% of employees in remote-capable roles use AI at least a few times a week, including 26% daily, against 22% in jobs that require on-site work (Q3 2026). Gallup
  14. 52% of frequent AI users spend less time searching for or gathering information, and 47% spend less time writing or drafting (Q3 2026). Gallup
  15. 29% of employees in AI-adopting organizations say AI has raised expectations for daily output (Q3 2026). Gallup
  16. Writing and editing is the most common workplace use, at 51% of employee AI users, ahead of search or research (49%) (Q2 2026). Gallup

What this shows: Worker-reported use (56% have used AI for at least one task) runs far ahead of firm-reported adoption (about 20%), which suggests a lot of use happens without a formal company rollout. Daily use is still a minority habit, between 16% (Gallup) and 24% of AI users (Census).

Consumers and chatbots

Population surveys answer a different question: how many people, at home or at work, use generative AI or chatbots at all. Pew asks U.S. adults, Eurostat asks EU residents aged 16 to 74, and Microsoft estimates usage from its own device telemetry, adjusted to each country's working-age population.

Pew changed its 2026 question to include other chatbots, so the jump from 2025 is partly a wording effect.

Generative AI and chatbot use among adults

Share of adults using AI chatbots or generative AI tools, latest official or survey estimate

U.S. adults: any chatbot49%U.S. adults: ChatGPT44%EU 16-74: gen AI32.7%Denmark 16-7448.4%Romania 16-7417.8%World 15-64 (Microsoft)17.8%UAE 15-64 (Microsoft)70.1%
Pew: U.S. adults, Feb 2026. Eurostat: ages 16-74, 2025. Microsoft: working-age 15-64, Q1 2026, telemetry estimate. Chart by ReviewRamble. Data: Pew Research Center, Eurostat, Microsoft.
  1. About half of U.S. adults (49%) now use AI chatbots, up from a third (33%) in 2024; Pew notes the 2024 wording differed. Pew Research Center
  2. 44% of U.S. adults report using ChatGPT, up from 34% in 2025 and more than double the 2023 share. Pew Research Center
  3. Adults under 50 are about twice as likely as those 50 and older to use ChatGPT (57% vs. 28%) (Feb 2026). Pew Research Center
  4. 13% of U.S. adults use chatbots to get news (Feb 2026). Pew Research Center
  5. Men are more likely than women to use chatbots daily (27% vs. 20%), even though overall use is now similar (Feb 2026). Pew Research Center
  6. 22% of U.S. men have used Microsoft Copilot, against 13% of women (Feb 2026). Pew Research Center
  7. Only 5% of U.S. adults who do not use chatbots think it is highly likely they will start within a year; another 16% say somewhat likely (Feb 2026). Pew Research Center
  8. 8% of U.S. adults say they are using AI chatbots to find election-related information for the 2026 election (Oct 12-18, 2026). Pew Research Center
  9. 32.7% of people aged 16-74 in the EU used generative AI tools in 2025 (use in the 3 months before the survey). Eurostat
  10. 25.1% of EU residents aged 16-74 used generative AI for personal purposes in 2025, against 15.1% for work and 9.4% for formal education. Eurostat
  11. Denmark had the EU's highest generative AI use, 48.4% of people aged 16-74, followed by Estonia (46.6%) and Malta (46.5%) in 2025. Eurostat
  12. Romania had the EU's lowest generative AI use, 17.8%, followed by Italy (19.9%) and Bulgaria (22.5%) in 2025. Eurostat
  13. 64% of EU residents aged 16-24 used generative AI in 2025, against 7% of those aged 65-74. Eurostat
  14. 49% of EU residents with a high level of formal education used generative AI in 2025, against 26% with a medium level. Eurostat
  15. 64% of EU non-users said the main reason was that they did not need generative AI; 14% were unsure how to use it (2025). Eurostat
  16. 17.8% of the world's working-age population used AI in Q1 2026, up 1.5 percentage points from 16.3%, per Microsoft's telemetry-based estimate. Microsoft
  17. The UAE led Microsoft's national AI diffusion leaderboard at 70.1% of its working-age population in Q1 2026. Microsoft
  18. The U.S. ranked 21st on Microsoft's leaderboard at 31.3%, up from 24th (Q1 2026). Microsoft
  19. AI usage reached 27.5% of the working-age population in the Global North against 15.4% in the Global South (Q1 2026). Microsoft

What this shows: Age is the strongest predictor in both the U.S. and EU data: under-30s and 16-24s use these tools at roughly two to nine times the rate of the oldest groups. Most non-users say they simply do not need it, rather than citing cost or access. That matches the large share of non-users who say they are unlikely to start within a year.

Developers and AI coding tools

Developer surveys show the highest use of any group, but they are opt-in samples of engaged developers (Stack Overflow) or targeted panels (JetBrains), not random samples of all programmers. GitHub's Octoverse is platform data from GitHub itself.

How much developers trust AI output

Stack Overflow 2025 respondents, trust in the accuracy of AI tools

Highly trust3.1%Somewhat trust29.6%Somewhat distrust26.1%Highly distrust19.6%
Stack Overflow Developer Survey 2025, n=33,244; remaining share neither trusts nor distrusts. Chart by ReviewRamble. Data: Stack Overflow.
  1. 84% of developers use or plan to use AI tools in their development process, up from 76% in 2024 (Stack Overflow 2025). Stack Overflow
  2. 51% of professional developers use AI tools daily (Stack Overflow 2025). Stack Overflow
  3. 46% of developers distrust the accuracy of AI tools, while 33% trust it, and only 3% highly trust the output (Stack Overflow 2025). Stack Overflow
  4. Positive sentiment toward AI tools fell to 60% in 2025, from over 70% in 2023 and 2024 (Stack Overflow). Stack Overflow
  5. 66% of developers cite "AI solutions that are almost right, but not quite" as their biggest frustration; 45% say debugging AI code takes longer (2025). Stack Overflow
  6. 72% of developers say they are not vibe coding, and another 5% reject it emphatically (Stack Overflow 2025). Stack Overflow
  7. ChatGPT (82%) and GitHub Copilot (68%) lead out-of-the-box AI assistants among developers who use them (Stack Overflow 2025). Stack Overflow
  8. 87% of developers are concerned about the accuracy of AI agents and 81% about data security and privacy (Stack Overflow 2025). Stack Overflow
  9. Coding assistants and agents are now the top AI use case, at 66% of developers, ahead of general-purpose chatbots (63%) (Stack Overflow 2026). Stack Overflow
  10. 31% of developers who use AI daily use it for more than four hours a day (Stack Overflow 2026, n=10,926). Stack Overflow
  11. Claude Code (66%) and GitHub Copilot (59%) are the most used coding agents in the 2026 Stack Overflow survey. Stack Overflow
  12. Only 6% of developers switched AI tools because of cost; 25% switched for better output quality (Stack Overflow 2026). Stack Overflow
  13. 17% of developers who avoid AI cite fear of losing skills, the top reason; 41% do not avoid AI at all (Stack Overflow 2026, n=13,857). Stack Overflow
  14. 90% of developers regularly used at least one AI tool at work for coding and development in January 2026 (JetBrains). JetBrains Research
  15. 74% of developers had adopted specialized AI coding tools such as assistants, editors and agents, not just chatbots (January 2026, JetBrains). JetBrains Research
  16. GitHub Copilot is known to 76% of developers and used at work by 29%, the highest of any AI coding tool (January 2026, JetBrains). JetBrains Research
  17. Cursor and Claude Code are each used at work by 18% of developers worldwide (January 2026, JetBrains). JetBrains Research
  18. About 22% of developers have over 80% of their code generated by agents (May-July 2026, JetBrains). JetBrains Research
  19. On average, 58% of Cursor users' code is agent-generated (May-July 2026, JetBrains). JetBrains Research
  20. More than 36 million developers joined GitHub in a year, bringing the total to over 180 million (Sept 2024 to Aug 2025). GitHub
  21. Nearly 80% of new developers on GitHub use Copilot in their first week (Sept 2024 to Aug 2025). GitHub
  22. More than 1.1 million public repositories use an LLM SDK, with 693,867 created in the past 12 months (+178% year on year, Aug 2025 vs Aug 2024). GitHub

What this shows: Use and trust have moved apart. Nine in ten developers now use AI at work and coding agents are the main use case, but more developers distrust AI accuracy than trust it, and the most common frustration is code that is "almost right". For tool choices, see our reviews of coding agents and AI code editors.

Investment and the model ecosystem

Investment and model counts show where adoption is supplied from. The AI Index reports private investment from commercial deal data, which counts private funding rounds and leaves out government spending. Model counts come from Epoch AI's notable-models dataset, which lists models that pass its notability criteria. That is a curated list, not a census of every model.

Epoch counts below were calculated by ReviewRamble from notable_ai_models.csv downloaded on 2026-11-18. Epoch adds models after the fact, so counts for recent years tend to rise.

Notable AI models published per year

Count of models in Epoch AI's notable-models dataset, by publication year

20196820205020217920229120231192024992025111
Calculated by ReviewRamble from Epoch AI notable_ai_models.csv (CC BY 4.0), downloaded 2026-11-18. Chart by ReviewRamble. Data: Epoch AI.
  1. U.S. private AI investment reached $285.9 billion in 2025, more than 23 times China's $12.4 billion (2026 AI Index). Stanford HAI
  2. 1,953 newly funded AI companies were founded in the U.S. in 2025, more than 10 times the next closest country (2026 AI Index). Stanford HAI
  3. Industry produced over 90% of notable frontier models in 2025 (2026 AI Index). Stanford HAI
  4. Top-model performance on the SWE-bench Verified coding benchmark rose from 60% to near 100% in a single year (2026 AI Index). Stanford HAI
  5. AI agents went from 12% to about 66% task success on OSWorld, a benchmark of real computer tasks (2026 AI Index). Stanford HAI
  6. 111 notable AI models were published in 2025, up from 99 in 2024 but below 2023's 119 (Epoch AI data, calculated by ReviewRamble). Epoch AI
  7. 70 of 2025's 111 notable models involved a U.S.-based organization, and 43 a China-based one (models can list more than one country; Epoch AI data, calculated by ReviewRamble). Epoch AI
  8. 101 of 2025's 111 notable models came solely from industry organizations, against 66 of 119 in 2023 (Epoch AI data, calculated by ReviewRamble). Epoch AI
  9. 47 of 2025's 111 notable models (42%) were released with open weights, counting unrestricted, restricted-use and non-commercial licences (Epoch AI data, calculated by ReviewRamble). Epoch AI
  10. 47 of 2025's 111 notable models were offered through an API, the most common access route, while 8 were unreleased (Epoch AI data, calculated by ReviewRamble). Epoch AI
  11. 36 of 2023's 119 notable models were never released, against 8 of 111 in 2025 (Epoch AI data, calculated by ReviewRamble). Epoch AI

What this shows: Model output and investment are concentrated, but the openness of notable models has shifted. In our count, 42% of 2025's notable models were released with open weights. Browse current models and prices on our models directory and AI model pricing tracker.

Trust, risk and public attitudes

Attitude surveys measure how people feel about AI, not what it does. Incident counts measure what gets reported, and reporting rises with attention, so a higher count does not necessarily mean AI became less safe.

How U.S. adults see AI

Share of U.S. adults, February 2026

AI advancing too quickly63%Personal info less secure71%Little/no confidence in gov. regulation67%Negative impact on society40%Negative impact on them31%AI advancing too slowly2%
Pew Research Center survey of 5,119 U.S. adults, Feb 17-23, 2026. Chart by ReviewRamble. Data: Pew Research Center.
  1. Documented AI incidents rose to 362 in 2025, up from 233 in 2024 (2026 AI Index). Stanford HAI
  2. 73% of AI experts expect a positive impact on how people do their jobs, compared with 23% of the public, a 50-point gap (2026 AI Index). Stanford HAI
  3. Only 6% of teachers say their school's AI policies are clear, though over 80% of U.S. high school and college students use AI for school (2026 AI Index). Stanford HAI
  4. 63% of U.S. adults say AI is advancing too quickly; only 2% say too slowly (Feb 2026). Pew Research Center
  5. 40% of U.S. adults say AI will have a negative impact on society over the next 20 years; far fewer expect a positive impact (Feb 2026). Pew Research Center
  6. 31% of U.S. adults expect AI to have a negative effect on them personally over the next two decades, against about a quarter who expect a positive one (Feb 2026). Pew Research Center
  7. About 71% of U.S. adults predict AI will make their personal information less secure; just 3% say more secure (Feb 2026). Pew Research Center
  8. 67% of Americans have little to no confidence in the U.S. government to regulate AI effectively, up from 62% in 2024 (Feb 2026). Pew Research Center
  9. 74% of Democrats and 61% of Republicans lack confidence in government AI regulation (including leaners, Feb 2026). Pew Research Center
  10. 68% of U.S. women and 58% of men say AI is moving too fast (Feb 2026). Pew Research Center
  11. 33% of U.S. women expect AI to have a negative personal impact over 20 years, twice the 17% expecting a positive one (Feb 2026). Pew Research Center

What this shows: The public is more worried than the people building AI: experts are three times as likely as the public to expect AI to improve how people do their jobs. Most U.S. adults want AI to slow down, and confidence that the government can regulate it is falling.

Methodology and definitions

  • Source selection. We used official statistics offices first (U.S. Census Bureau, Eurostat, ONS), then probability or large-panel surveys (Pew Research Center, Gallup), annual developer surveys (Stack Overflow, JetBrains), platform data published by its owner (GitHub, Microsoft) and research compilations with published methods (Stanford AI Index, Epoch AI). Vendor marketing surveys and statistics roundups were excluded.
  • Periods. The 2026 in the title is the year of review. Every figure keeps its own measurement period, and some are from 2025.
  • Definitions differ. "Using AI" means use in the last two weeks for U.S. firms (BTOS), during the year for EU enterprises, and use of at least one listed technology for UK firms (BICS). Population surveys ask about chatbots or generative AI. Census changed its BTOS wording in November 2025, and Pew changed its chatbot question in 2026.
  • Our calculations. Epoch AI counts are ReviewRamble tallies of the public notable_ai_models.csv (downloaded 2026-11-18), with the filter stated for each figure. No other figure is our own research.
  • Verification. Each source page was downloaded on 2026-11-18, and each figure was matched to the text of its source. The supporting excerpt and its location are in the downloadable CSV.
  • Not used. Statistics Canada and the St. Louis Fed publish relevant data, but their pages could not be retrieved for verification on the review date, so they are excluded.

Corrections and updates

We recheck links monthly and update figures when a source publishes a new release (BTOS every two weeks, Eurostat each December, Pew each year, AI Index each spring). Old figures are replaced, not piled up. Report an error to support@reviewramble.com.

Download all figures with sources (CSV)

Using these figures

You are welcome to quote or chart these figures. Please cite the original source linked at each figure, and credit ReviewRamble with a link to this page if you use our charts or comparisons.

ReviewRamble, “AI Adoption Statistics 2026”, Nov 18, 2026, https://www.reviewramble.com/research/ai-adoption-statistics

Original sources

Further reading