Short answer: On the one like-for-like yardstick available, Ireland leads these five countries, followed by the UK, Australia, Canada and the US. Official business surveys define “AI use” differently in each country and put adoption at roughly 12% to 35% of firms. Almost everywhere, people use AI more than their employers do, and most business use is shallow.
Ask four credible sources how many Australian businesses use AI and you’ll get four answers: 12%, 44%, 72% and 88%.
They’re all real. The Australian Bureau of Statistics counts 12%. The government’s own SME tracker says 44%. A domain-registry survey found 72% of small businesses had used AI. An accountants’ body reports 88% using it in some form. Nobody is wrong. Each one is answering a different question, and that’s the key to reading every AI adoption statistic you’ll see for the US, Canada, the UK, Ireland and Australia.
This guide sorts out which numbers you can compare, which you can’t, and what the picture looks like once you do the sorting.
Why AI adoption statistics never match
Every national statistics office measures something slightly different:
- United States (Census Bureau): employer firms using AI “in any business function” in the last two weeks. About 18% at the end of 2025, and just under 20% by spring 2026.
- Canada (Statistics Canada): businesses that used AI to produce goods or deliver services in the past 12 months. 19.2% in Q2 2026.
- UK (ONS): businesses with 10 or more employees using at least one listed AI technology. About 35% in June 2026.
- Ireland (CSO, Eurostat-style): enterprises with 10 or more employees using at least one AI technology. About 20% in 2025.
- Australia (ABS): employing businesses that ticked AI on a list of technologies used in 2024–25. 12%, and the ABS says the question isn’t designed to measure how intensively it’s used.
The wording matters more than you’d think. In November 2025, the US Census Bureau broadened its question from AI used “in producing goods or services” to AI used in “any business function.” The measured rate jumped from about 10% to about 18% without a single firm changing behaviour (Federal Reserve).
Who you ask matters too. The same Fed note reports four defensible US figures: 18% of firms (counting each firm once), 41% of workers using generative AI at work, 54% of the labour force working at firms that use LLMs, and 78% working at firms that use AI in some form, as reported by executives. Earlier Fed research reviewed 16 surveys and found work-related AI adoption estimates ranging from about 5% to 40% as of mid-2024.
The practical rule: compare a country with its own past, not with another country’s headline.
The one like-for-like ranking
Microsoft’s AI Economy Institute tracks the share of people aged 15 to 64 who used a generative AI product, using aggregated telemetry adjusted for device market share, internet penetration and population (Microsoft, Q2 2026). The same method across 147 economies is what makes it comparable. June 2026 results:
- Ireland: 49.9% (third in the world)
- United Kingdom: 43.7%
- Australia: 40.7%
- Canada: 38.5%
- United States: 33.0% (21st)
For context, the world average was 18.8%, and the UAE (73.3%) and Singapore (64.3%) led. The order of these five countries hasn’t changed since 2025, and all of them are growing. Between the first half of 2025 and June 2026, Ireland rose from 41.7% to 49.9%, the UK from 36.4% to 43.7%, Australia from 34.5% to 40.7%, Canada from 33.5% to 38.5% and the US from 26.3% to 33.0% (Microsoft, H2 2025).
Two caveats. It tracks Microsoft-visible generative AI use, not intensity, and Microsoft says it will widen the tools it counts in the next report. So treat it as a trend lens and a rough ordering, not a precise league table.
AI adoption in Ireland: ahead with people, mid-pack with firms
Ireland tops the ranking, and household data agree. Eurostat-based figures put 44.9% of Irish people aged 16 to 74 using generative AI in 2025, against an EU average of 32.7% (EU Digital Skills and Jobs Platform). The CSO found 65% of 16–29-year-olds had used generative AI tools, compared with 15% of 60–74-year-olds and 6% of people aged 75 and over (CSO).
Businesses are less far along. About 20% of enterprises used AI in 2025, up from 15% in 2024 and 8% in 2023. That matches the EU average of 20% and trails Denmark (42%) and Finland (37.8%) (CSO). Size explains most of the gap: roughly 58% of large enterprises use AI, against about 17% of small ones (CSO release). Ireland’s people are racing ahead while its smaller firms are still at the starting line.
AI adoption in the UK: tripled in under three years, but thin
The ONS reports that around 35% of UK businesses with 10 or more employees used at least one AI technology in June 2026, up from around 12% in late 2023 (ONS). Reporting on the same release puts large firms at 49% and the smallest at 28%, with information and communication at 58% and construction at 13%. Large language models were the most used technology (18%), followed by visual content creation (16%).
The more interesting finding is how shallow it is. Adopters use about 1.6 AI technologies each, up from 1.4 in 2023, and only about 10% call their use extensive. The ONS reads that as limited transformative impact for most adopters. Meanwhile, 55% of employed and self-employed people told the ONS they use AI for work or education, which means workers are well ahead of their employers, and just 11% of businesses with 10 or more staff say most of their workforce has had AI training (Onrec, citing ONS).
The mood is cautious. In Ipsos’s 2026 survey, 62% of Britons said AI products make them nervous and 29% said excited, against 32-country averages of 50% and 51% (Ipsos).
AI adoption in Australia: the widest range of numbers
The ABS found 12% of employing businesses used AI in 2024–25, up from 1% in 2021–22. Adoption was 20% among innovation-active businesses and 6% among those that weren’t (ABS). As reported, the figure is around 35% for large businesses and about 11% for small ones. The ABS plans to add more AI questions to its 2026–27 survey.
The other numbers come from different questions. The National AI Centre’s monthly tracker, which asks 400 SME decision-makers how they describe their own adoption, reported 44% in February 2026. auDA’s 2026 survey found 72% of small businesses and 61% of consumers had used AI (auDA). An ANU and Google survey found 48.6% of adults had used generative AI (ANU). CPA Australia’s October 2026 report found 88% of businesses using AI in some form, but only 10% with it widely embedded, against 14% across the markets it compared.
Trust lags use. A 2025 global study found 50% of Australians use AI regularly but only 36% are willing to trust it (KPMG). Policy is light-touch: the December 2025 National AI Plan relies on existing laws and a new AI Safety Institute rather than a standalone AI Act (ABC).
AI adoption in Canada: tripled in two years
Statistics Canada found 19.2% of businesses used AI to produce goods or deliver services in the 12 months to Q2 2026, up from 12.2% a year earlier and 6.1% in 2024 (Statistics Canada). Information and cultural industries led at 42.3%, finance and insurance followed at 40.4%, and professional services at 32.4%. Primary agriculture sat at just 4.5%. The top uses were data analytics (36.6%), text analytics (34.5%) and virtual agents or chatbots (28.2%).
The barriers are telling. Two in five businesses (40.0%) said AI isn’t relevant to what they produce, 23.0% reported no barriers, 13.4% cited cybersecurity or privacy concerns and 10.6% cited cost. Intent is rising: in Q3 2026, 25.2% of businesses said they plan to use AI over the next 12 months (Statistics Canada).
People and workers are ahead. In March 2026, 35.9% of Canadian workers had used generative AI tools in their main job over the previous 12 months, and 41.6% had used at least one AI or automation technology (Investing.com, reporting StatCan). CIRA’s 2026 survey found 46% of Canadians used a generative AI tool in the past year, up from 16% in 2024, while 56% were concerned about it (CIRA). Note that Canada’s 12-month window is far more generous than the US two-week window, so don’t compare 19.2% with 18% directly.
AI adoption in the US: the biggest gap between people and firms
The Census Bureau’s biweekly survey showed about 18% of firms using AI at the end of 2025 under the new question, and just under 20% in the first months of 2026. Over 20% expected to use it within six months (Minneapolis Fed). Less than 10% of businesses in agriculture, transportation, accommodation and food services, and construction reported AI use. Professional services (about 33%) and finance (about 30%) led in late 2025.
Firms look modest, but workers and consumers don’t. About 41% of the US workforce used generative AI for work in November 2025, and about 12% did so daily (Federal Reserve). Pew found 49% of US adults use AI chatbots (up from 33% in 2024), with 24% using them daily and 66% of 18–29-year-olds on board (Pew). Count by employment instead of by firm and the picture changes: a Census-based NBER paper found 18% of firms but 32% of employment at firms using AI, with 50–60% among very large firms in knowledge-intensive sectors (NBER).
The US also ranks lowest of the five on the Microsoft measure, 21st globally. That says more about the breadth of everyday use than about the strength of the AI industry.
Five patterns that show up in all five countries
1. People are ahead of employers. UK workers: 55% use AI for work or study versus 35% of firms. US workers: 41% versus about 18% of firms. Australian consumers: 54% use AI for work at least occasionally, versus the ABS’s 12% of businesses. Definitions differ, but the direction doesn’t. Expect plenty of unofficial, unmanaged use at work.
2. Size matters. Large firms adopt roughly two to four times as often as small ones: UK 49% versus 28%, Ireland about 58% versus 17%, Australia about 35% versus 11%, US 37% of 250+ employee firms versus under 20% of the tiniest. Canada is the exception, with 19.9% of one-to-four-employee firms close to the 19.2% average.
3. Information, finance and professional services lead. The UK’s 58% in information and communication, Canada’s 42.3% in information and culture and the US’s one-third in professional services tell the same story. Construction, agriculture and hospitality are at the other end everywhere.
4. Wide, not deep. The UK has about 10% extensive use among adopters, Australia about 10% embedded, and in the US a Census-based analysis found 57% of AI-using firms deploy it in three or fewer functions. Most firms are using chatbots for drafting and research, not rebuilding workflows.
5. A nervous public. Ipsos’s 2026 survey of 32 countries found 62% (Great Britain) to 67% (Australia and Canada) saying AI products make them nervous, against a 32-country average of 50%. The US and Ireland sat at 64% and 63% as reported. People are using AI while staying wary of it, which is why disclosure and trust show up in nearly every national strategy and survey.
What the numbers can’t tell you yet
- There’s no shared definition. The OECD, Eurostat and national offices are converging, but a cross-country business league table doesn’t exist yet.
- Survey wording can break trends. The US jump from 10% to 18% shows how a question change can masquerade as growth.
- Self-reports are fuzzy. “Used AI” can mean a one-off chatbot query or a governed production system.
- Telemetry has blind spots. Microsoft’s data captures what Microsoft can see, and its coverage is changing.
- Use isn’t value. None of these surveys shows whether AI is paying off, and the ONS’s own reading is that impact is limited for most adopters so far.
Trends within one country are your most reliable signal. Australia’s numbers should get more comparable once the ABS adds AI questions in 2026–27.
What this means for your business
- Benchmark against your own size and sector. A 12-person construction firm in the UK lives in a 13% world, not a 35% one.
- Treat chatbot use as step one. Pick two or three workflows, train the people who run them, and measure results. Only 11% of UK firms say most of their staff are trained, and 40% of Canadian firms say AI isn’t relevant to what they do.
- Assume your employees are already using AI. Give them approved tools and clear rules before shadow use does it for you.
- Plan for a wary audience. With six in ten people nervous in all five countries, tell customers when and how you use AI.
- Re-benchmark twice a year. Canada tripled in two years and the UK in under three.
The bottom line
AI adoption across these five countries is broad, fast and shallow. Ireland’s people lead, the UK’s businesses have moved fastest on paper, Canada has tripled in two years, Australia has the widest spread of statistics, and the US has the biggest gap between worker use and firm use. All five are growing, and all five publics are uneasy.
Before you quote any adoption figure, check three things: who was asked, what counts as use, and over what time window. Do that, and the numbers stop contradicting each other and start telling you something.
