← All articles

How Many Businesses Really Use AI? 16 Sources, and Why They Disagree

This analysis draws on data from 16 credible sources, including government statistics, central bank research, academic studies and industry reports.

Statistics Canada counted 19.2% of Canadian businesses using AI to produce goods or deliver services in the year to spring 2026.1 The US Census Bureau put the American figure at 19.8% in May 2026.4 Eurostat measured 19.95% across the European Union in 2025.8 Three statistics agencies, three different questions, and the same answer: about one business in five.

Ask executives instead and the number jumps. A survey of almost 6,000 CFOs, CEOs and executives in four countries found about 70% of firms actively using AI.7 McKinsey's 2026 survey has nearly nine in ten respondents reporting regular AI use in at least one business function.13

Both sets of numbers are measured carefully. They are counting different things. And under every one of them, the business function where AI shows up first is the same: marketing and sales.

We'll explore:

The official count: about one business in five

The national statistics agencies survey large, random samples of businesses and weight the answers to represent every firm in the economy. On that basis, adoption is real, rising fast, and still a minority.

In Canada, the share of businesses using AI has tripled in two years, from 6.1% in the second quarter of 2024 to 12.2% in 2025 and 19.2% in 2026.1,2 The 2026 figure comes from 9,251 responding businesses.1 Two in five Canadian businesses, 40.0%, told Statistics Canada that AI is not relevant to them at all.1

The US Census Bureau's biweekly survey shows AI use between 17% and 20% of firms from December 2025 to May 2026, with 20% to 23% expecting to use it within six months.4 A Census working paper drawing on more than 117,000 firms put the rate at 18% for November 2025 to January 2026.5

Europe is a little ahead on a narrower base. Eurostat counts only enterprises with 10 or more employees, and 19.95% of them used AI in 2025, up 6.47 percentage points in a year.8 The spread between countries is wide, from 42.03% in Denmark to 5.21% in Romania.8 Statistics Sweden recorded 35.0% of Swedish enterprises using AI in 2025, up from 25.1% in 2024 and 10.4% in 2023.10

The UK is the highest of the official counts. The Office for National Statistics found 30% of all UK businesses using at least one AI technology in September 2026, up from 9% in September 2023.9 Among UK firms with 10 or more employees the rate was 38%.9

Figure 1: Share of Businesses Using AI, Official Statistics (%)

Source: 1, 9, 10. Statistics Canada asks about AI used to produce goods or deliver services; Statistics Sweden covers enterprises with 10 or more employees; ONS covers all UK businesses.

The direction is the same everywhere. Every official series in this set is higher now than at its first reading. The levels differ partly because the questions differ: Statistics Canada asks about AI used to produce goods or deliver services, while the US Census Bureau has asked since November 2025 about AI used in any business function.1,4

Why executive surveys report 70% and more

A Federal Reserve Board note published in April 2026 set the competing American numbers side by side. The Census survey showed about 18% of firms using AI at the end of 2025. The Atlanta Fed's Survey of Business Uncertainty showed about 78% of the workforce employed at firms using AI.6 Both were right.

The gap is mostly about who gets counted. The Census survey counts every firm once, and about 57% of US firms have fewer than five employees.6 Firms with one to 49 employees make up 89.6% of the Census sample but only 38.2% of the uncertainty survey's sample, while firms with 250 or more employees are 2.2% of one and 27.4% of the other.6 Weight by employment instead of by firm, and the Census paper's own 18% becomes 32%.5

The executive surveys sit at the far end of that scale. The four-country survey of almost 6,000 executives found about 70% of firms using AI, and over two-thirds of top executives using it regularly, for an average of 1.5 hours a week.7 McKinsey's 2026 survey of 1,719 participants reported nearly nine in ten using AI regularly in at least one function, and 36% of its respondents work at organizations with more than $1 billion in annual revenue.13

The Bank of Canada found both numbers inside one survey. In its December 2025 Business Leaders' Pulse, 70% of respondents used AI for at least some time in a typical work week.11 Only 8% of the same businesses reported significant use of AI in their operations, with 21% reporting moderate use and 29% infrequent use.11 The bank's authors note that Canadian estimates range from about 15% to over 90% depending on the survey.11

A survey of owners lands in between. BDC found 30% of Canadian small and medium-sized businesses using generative AI in a February 2026 survey of 1,500 owners and decision makers, a non-probability sample that BDC says cannot be projected to the whole population.12

Figure 2: Reported Business AI Adoption, by What Each Survey Measures (%)

Source: 1, 4, 5, 6, 7, 8, 9, 11, 12. Grey bars measure individuals or weight by employment rather than counting each firm once.

Size decides more than anything else

Every source that breaks adoption down by company size finds the same gradient. In the EU in 2025, 17% of small enterprises used AI, against 30.36% of medium-sized ones and 55.03% of large ones.8 In Sweden the same three bands read 30.8%, 49.6% and 71.9%.10

In the US, 37% of firms with 250 or more employees used AI in spring 2026, and 32% of firms with 100 to 249.4 Very large firms in information, professional services and finance reached use rates of 50% to 60%.5 In Canada, 27.8% of businesses with 100 or more employees used AI, against 19.9% of those with one to four.1 Small Canadian firms were also nearly twice as likely to call AI irrelevant: 41.4% of businesses with one to four employees, against 21.3% of those with 100 or more.1

Figure 3: AI Use by Enterprise Size, EU and Sweden, 2025 (%)

Source: 8, 10.

Industry matters almost as much. Information and cultural industries led Canada at 42.3%, followed by finance and insurance at 40.4% and professional, scientific and technical services at 32.4%.1 Construction sat at 9.2% and agriculture at 4.5%.1 In the EU, 62.52% of information and communication enterprises used AI.8 In the UK the same sector reached 60%.9 Urban Canadian businesses used AI at more than twice the rural rate, 21.0% against 9.9%.1

Marketing is where AI lands first

The surveys disagree about how many businesses use AI. They agree about where it goes once a business starts.

The Census Bureau's working paper asked AI-using firms which of 15 business functions they used it in. Sales and marketing came first, at 52% of AI-using firms, ahead of strategy and business development at 45%, information technology at 41% and research and development at 40%.5 Sales and marketing was also among the functions where firms expected the largest increases over the next six months.5

Figure 4: Most Common Business Functions for AI Among US AI-Using Firms (%)

Source: 5. Share of firms using AI in at least one business function, November 2025 to January 2026.

Europe reports the same order. Marketing or sales was the most common purpose for AI in 2025, cited by 34.70% of AI-using EU enterprises, ahead of business administration and management at 31.05%.8 In accommodation the share reached 58.82%, and in retail trade 48.18%.8 In Sweden, marketing or sales has been the top purpose for three years, rising from 28.1% of AI users in 2023 to 39.9% in 2024 and 41.7% in 2025.10

McKinsey's executives point the same way on results. Respondents most often attribute revenue gains from AI to its use in marketing and sales.13

Canada's data shows a more mixed picture. Statistics Canada tracks applications rather than functions, and marketing automation rose from 15.2% of AI-using businesses in 2024 to 23.1% in 2025.2 In 2026 it fell to 19.8%, as data analytics rose to 36.6% and large language models to 24.8%.1 Among Canadian businesses planning to adopt AI, 25.6% named marketing automation as a planned use in 2025, up from 19.4% a year earlier.3 The youngest firms lean on it hardest: 52.0% of AI-using businesses two years old or less used marketing automation in 2026, against 14.5% to 22.2% for older firms.1

Figure 5: Marketing Use as a Share of AI-Using Businesses (%)

Source: 1, 2, 10. Two different measures: Sweden asks for the purpose of AI use, Canada for the type of AI application.

Inside the marketing department

Marketing leaders are now budgeting for AI as a line of its own. Gartner's 2026 CMO Spend Survey of 401 marketing leaders found an average of 15.3% of marketing budgets allocated to AI initiatives.14 Marketing budgets themselves barely moved, at 7.8% of company revenue in 2026 against 7.7% in 2025.14

Readiness lags the spending. Seventy percent of the CMOs Gartner surveyed called becoming an AI leader a critical goal for 2026, and only 30% reported mature or fully developed AI readiness.14 The more AI-ready marketing organizations put 21.3% of their budgets into AI.14

Two years earlier the base was small. In The CMO Survey's 2024 edition, 292 marketing leaders reported generative AI in 7% of their marketing activities, and only 10% of companies were already using large language models in marketing.16 Another 51% were evaluating or piloting them.16

Individual use runs ahead of company use here too. As of May 2026, 52% of US employees used AI at work a few times a year or more, 30% a few times a week or more and 15% daily, according to Gallup.15 Only 47% said their organization had integrated AI tools, and 25% said it had communicated a clear plan.15 A separate survey cited by the Federal Reserve put work-related generative AI use at about 41% of the US workforce in November 2025.6

What AI has changed so far

On jobs, very little yet. Among Canadian businesses using AI in 2025, 89.4% reported no change in employment from it, 6.3% a decrease and 4.3% an increase.2 In the UK, 6% of AI-using firms with 10 or more employees reported reduced headcount, and fewer than 1% increased headcount.9 The Census paper found AI-related employment decreases in 2% of US firms, and 66% of users relying on AI only to augment tasks.5

On productivity, executives report little so far and expect more. More than 80% of firms in the four-country executive survey reported no impact of AI on employment or productivity over the previous three years.7 The same firms forecast AI would raise productivity by 1.4%, increase output by 0.8% and cut employment by 0.7% over the next three years.7 In Canada, 18% of Business Leaders' Pulse firms expect to hire fewer staff because of AI over three years, against 9% who expect to hire more.11

On profit, the share seeing results has stopped growing. In McKinsey's 2026 survey, 37% of respondents attributed at least some EBIT impact to AI, essentially unchanged from 2025, even as the share scaling AI across the enterprise rose from 38% to 44%.13 High performers, who attribute at least 5% of EBIT to AI, stayed at about 6% of respondents.13 BDC reported that Canadian SMEs using generative AI were 24% more productive than those that do not.12

The barriers are mostly about relevance and know-how. Among Canadian businesses with no plans to use AI, 78.1% said it was not relevant to their goods or services in 2025.3 Among all Canadian businesses in 2026, cybersecurity or privacy was the most-cited barrier at 13.4%, ahead of cost at 10.6%.1 In the EU, 70.89% of enterprises that considered AI but did not adopt it cited a lack of relevant expertise.8

What the data does and does not show

Eleven of these 16 sources are published by government agencies or central banks, and the official counts among them agree closely: roughly one business in five in Canada, the US and the EU, and three in 10 in the UK.1,4,8,9 They count every firm equally, which means they mostly count very small firms.6 The executive surveys over-represent large firms and often measure personal use, which is why they report 70% and more.6,7,11

The definitions also move. The Census Bureau changed its question in November 2025, so its earlier and later figures are not directly comparable.4 Statistics Canada measures AI used to produce goods or deliver services, a narrower question than use anywhere in the business.1 BDC's figure covers generative AI only and comes from a non-probability panel.12 The CMO Survey figures are from 2024 and are used here only as a baseline.16

Within those limits, three findings hold across every source that measures them. Adoption is higher now than when each series began. Large firms adopt at higher rates than small ones in every source that splits by size. And among firms that use AI, marketing and sales is the most common place to use it.5,8,10

Sources

  1. Statistics Canada. Analysis on artificial intelligence use by businesses in Canada, second quarter of 2026. June 11, 2026.

  2. Statistics Canada. Analysis on artificial intelligence use by businesses in Canada, second quarter of 2025. June 16, 2025.

  3. Statistics Canada. Analysis on expected use of artificial intelligence by businesses in Canada, third quarter of 2025. September 11, 2025.

  4. US Census Bureau. Large Firms With at Least 20 Employees Biggest AI Users. May 26, 2026.

  5. US Census Bureau, Center for Economic Studies. The Microstructure of AI Diffusion. Working paper CES-WP-26-25, April 2026.

  6. Board of Governors of the Federal Reserve System. Monitoring AI Adoption in the U.S. Economy. FEDS Notes, April 3, 2026.

  7. Federal Reserve Bank of Atlanta. Firm Data on AI. Working Paper 2026-3, March 24, 2026.

  8. Eurostat. Use of artificial intelligence in enterprises. Statistics Explained, data extracted December 2025.

  9. Office for National Statistics. Business insights and impact on the UK economy: 8 October 2026. October 8, 2026.

  10. Statistics Sweden. Artificial intelligence in Sweden 2025. December 11, 2025.

  11. Bank of Canada. Survey Evidence on Firm AI Adoption and its Implications. Staff Analytical Paper 2026-22, June 2026.

  12. BDC. A $350B opportunity: Canada's next phase of growth to be driven by AI and digital technologies. June 3, 2026.

  13. McKinsey & Company. The state of AI in 2026: On the road to ROI. August 25, 2026.

  14. Gartner. 2026 CMO Spend Survey. May 11, 2026.

  15. Gallup. Indicator: Artificial Intelligence. Data as of May 2026, accessed October 26, 2026.

  16. Duke University Fuqua School of Business. The CMO Survey, 32nd edition. April 24, 2024.