AI adoption index
Who is using AI,
and what it is being used for.
Published figures from national statistics offices, the OECD and named industry surveys, charted. Every figure carries the population it is a share of and how it was collected, because nine different populations and three kinds of survey appear here, and several of them are routinely confused with each other.
What the data says
- 88% of organisations McKinsey surveyed report regular AI use in at least one business function, but only 25% of those Deloitte surveyed have moved 40% or more of their AI pilots into production.
- 54% of organisations on SHRM's HR panel have adopted no form of AI in HR and have no plans to do so this year. Where HR does use it, recruiting leads at 27%.
- 70.89% of EU enterprises that considered AI and did not adopt it gave lack of relevant expertise as a reason, the most cited of the eight reasons Eurostat lists.
- 64.0% of Dubai's working-age population used generative AI in the second half of 2025, against a global average of 16.3% across roughly 150 economies.
- 40% of US workers who have used AI chatbots for work say they were highly helpful in doing things more quickly. Among workers who use them at least a few times a month, 54% say so.
Every chart also exists as a table. The tables are always present for screen readers and crawlers; this control decides whether they are visible.
Where adoption actually is
How far adoption has gone, in the two industry surveys large enough to carry a global claim, and in the national statistics that count businesses one country at a time.
88% of the organisations McKinsey surveyed report regular AI use somewhere in the business. In Deloitte's separate survey, a quarter have moved 40% or more of their pilots into production.
88%
report regular AI use in at least one business function.
25%
have moved 40% or more of their AI pilots into production.
60%
of the workforce now has access to AI tools.
How this was measured
Three different industry surveys of organisations. Named on each figure
Industry survey (tier 3)
August to November 2025
Three figures from two separate surveys with separate samples and separate questions. They describe the same territory rather than the same respondents, so the 88% and the 25% are not two readings of one population.
| Category | Figure | What it measures | Basis | Year | Source |
|---|---|---|---|---|---|
| report regular AI use in at least one business function | 88% | report regular AI use in at least one business function | Organisations surveyed | November 2025 | McKinsey, The state of AI |
| have moved 40% or more of their AI pilots into production | 25% | have moved 40% or more of their AI pilots into production | Organisations surveyed | Fieldwork August to September 2025 | Deloitte, State of AI in the Enterprise |
| of the workforce now has access to AI tools | 60% | of the workforce now has access to AI tools | Organisations surveyed | Fieldwork August to September 2025 | Deloitte, State of AI in the Enterprise |
Where national statistics offices count businesses, adoption runs from 19.2% to 35%. Singapore is the one market that publishes large enterprises and SMEs separately, and they are 48 percentage points apart.
- United Kingdom35%
- Saudi Arabia33.1%
- United States19.8%
- Canada19.2%
- Singapore, larger enterprises62.5%
- Singapore, SMEs14.5%
Select any bar for what its figure means.
Five markets across four different populations. The rows are grouped by population and sorted inside each group rather than ranked across all of them, because a single ranked list would assert that the four are comparable. They are not.
How this was measured
Four different populations, grouped below and named on every row
National statistics (tier 1)
2024 to 2026
Saudi Arabia appears elsewhere on this page at 26.2%, as a share of the working-age population. Here it is 33.1%, a share of establishments. Those are not two versions of one number, and every row here names the population it is measured against.
| Category | Figure | What it measures | Basis | Year | Source |
|---|---|---|---|---|---|
| United Kingdom | 35% | of businesses with 10 or more employees used at least one AI technology | UK businesses with 10 or more employees | June 2026 | Office for National Statistics |
| Saudi Arabia | 33.1% | of establishments used AI technologies | Saudi establishments | 2025 | GASTAT, Establishments' ICT Access and Usage Statistics |
| United States | 19.8% | of businesses reported using AI in producing goods or services | US businesses of any size | May 2026 | US Census Bureau, Business Trends and Outlook Survey |
| Canada | 19.2% | of businesses used AI to produce goods or deliver services in the preceding twelve months | Canadian businesses of any size | Q2 2026 | Statistics Canada, Canadian Survey on Business Conditions |
| Singapore, larger enterprises | 62.5% | of larger enterprises had adopted AI | Singapore larger enterprises | 2024 | IMDA, Annual Survey on Infocomm Usage by Enterprises |
| Singapore, SMEs | 14.5% | of SMEs had adopted AI | Singapore SMEs | 2024 | IMDA, Annual Survey on Infocomm Usage by Enterprises |
What it's used for
Adoption is usually published by industry. These are the three sources that break it down by what the work actually is.
Marketing and sales is the most common purpose for AI among EU enterprises that already use it, at 34.70%. Logistics is the least common, at 6.08%.
- Marketing or sales34.70%
- Organisation of business administration processes or management31.05%
- Accounting, controlling or financial management23.11%
- Production processes20.76%
- ICT security19.77%
- Research and development or innovation activity19.05%
- Logistics6.08%
Select any bar for what its figure means.
How this was measured
Enterprises that already use AI, EU
National statistics (tier 1)
2025
| Category | Figure | What it measures | Basis | Year | Source |
|---|---|---|---|---|---|
| Marketing or sales | 34.70% | of AI-using enterprises use AI for this purpose | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Organisation of business administration processes or management | 31.05% | of AI-using enterprises use AI for this purpose | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Accounting, controlling or financial management | 23.11% | of AI-using enterprises use AI for this purpose | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Production processes | 20.76% | of AI-using enterprises use AI for this purpose | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| ICT security | 19.77% | of AI-using enterprises use AI for this purpose | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Research and development or innovation activity | 19.05% | of AI-using enterprises use AI for this purpose | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Logistics | 6.08% | of AI-using enterprises use AI for this purpose | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
Large enterprises are ahead of small ones on all three purposes Eurostat splits by size. The distance is widest in production processes: 19.02% of small enterprises against 33.46% of large.
Marketing
- 34.06%
- 35.67%
- 37.89%
Operations
- 19.02%
- 21.82%
- 33.46%
Finance
- 22.20%
- 24.21%
- 28.26%
How this was measured
Enterprises that already use AI, EU
National statistics (tier 1)
2025
| Category | Figure | What it measures | Basis | Year | Source |
|---|---|---|---|---|---|
| Marketing, Small | 34.06% | of AI-using enterprises of this size use AI for marketing or sales | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Marketing, Medium | 35.67% | of AI-using enterprises of this size use AI for marketing or sales | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Marketing, Large | 37.89% | of AI-using enterprises of this size use AI for marketing or sales | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Operations, Small | 19.02% | of AI-using enterprises of this size use AI in production processes | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Operations, Medium | 21.82% | of AI-using enterprises of this size use AI in production processes | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Operations, Large | 33.46% | of AI-using enterprises of this size use AI in production processes | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Finance, Small | 22.20% | of AI-using enterprises of this size use AI for accounting or financial management | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Finance, Medium | 24.21% | of AI-using enterprises of this size use AI for accounting or financial management | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Finance, Large | 28.26% | of AI-using enterprises of this size use AI for accounting or financial management | Enterprises that already use AI, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
71%
of organisations using AI in marketing and sales report revenue increases from it.
49%
of organisations using AI in service operations report cost savings from it.
4%
of UK businesses currently using AI reported that their overall workforce headcount had decreased as a result.
How this was measured
Three different populations: organisations already using AI in the named function, and UK businesses currently using AI. Named per figure in the table
Intergovernmental or peer-reviewed (tier 2)
2024 to December 2025
This measures workforce impact, not AI adoption within HR. It answers a different question and is shown here because it is the nearest measured statistic, not as a substitute.
| Category | Figure | What it measures | Basis | Year | Source |
|---|---|---|---|---|---|
| Marketing, Second source | 71% | of organisations using AI in marketing and sales report revenue increases from it | Organisations using AI in this function | 2024 | Stanford HAI, AI Index 2025 |
| Operations, Second source | 49% | of organisations using AI in service operations report cost savings from it | Organisations using AI in this function | 2024 | Stanford HAI, AI Index 2025 |
| People, Workforce impact, not HR adoption | 4% | of UK businesses currently using AI reported that their overall workforce headcount had decreased as a result | UK businesses currently using AI | 15 to 28 December 2025 | ONS, Business Insights and Conditions Survey |
54% of organisations on SHRM's HR panel have adopted no form of AI in HR, and have no plans to this year.
All organisations on the panel
- Recruiting27%
- HR technology21%
- Learning and development17%
- Employee experience14%
Select any bar for what its figure means.
Where HR does use AI, it is concentrated in recruiting. The four practice areas are each a share of all organisations on the panel and do not sum to the 39% that have adopted.
How this was measured
Organisations of 1,722 HR professionals on SHRM's opt-in Voice of Work panel
Industry survey (tier 3)
Fielded 5 to 23 December 2025
Three further practice areas are published as a ceiling rather than a value: inclusion and diversity, C-suite and board relations, and compliance, each at 2% or less. They are not drawn, because a bar would state a precision SHRM withheld. SHRM states the sample twice and differently: 1,908 in the analysis, 1,722 as the number who participated in the methodology. The figure used here is the participation count.
- Eurostat, Use of artificial intelligence in enterprises: seven purposes, none of them human resources or recruitment.
- OECD, The impact of AI on the workplace: human resources appears as an application area, with task-automation figures rather than an adoption rate.
- ONS: publishes AI adoption and workforce impact, not AI use within HR.
| Category | Figure | What it measures | Basis | Year | Source |
|---|---|---|---|---|---|
| Recruiting | 27% | of organisations use AI tools in this HR practice area | Organisations of 1,722 HR professionals on SHRM's opt-in Voice of Work panel | Fielded 5 to 23 December 2025 | SHRM, The State of AI in HR 2026 |
| HR technology | 21% | of organisations use AI tools in this HR practice area | Organisations of 1,722 HR professionals on SHRM's opt-in Voice of Work panel | Fielded 5 to 23 December 2025 | SHRM, The State of AI in HR 2026 |
| Learning and development | 17% | of organisations use AI tools in this HR practice area | Organisations of 1,722 HR professionals on SHRM's opt-in Voice of Work panel | Fielded 5 to 23 December 2025 | SHRM, The State of AI in HR 2026 |
| Employee experience | 14% | of organisations use AI tools in this HR practice area | Organisations of 1,722 HR professionals on SHRM's opt-in Voice of Work panel | Fielded 5 to 23 December 2025 | SHRM, The State of AI in HR 2026 |
| Category | Figure | What it measures | Basis | Year | Source |
|---|---|---|---|---|---|
| currently have AI adopted in their HR functions | 39% | currently have AI adopted in their HR functions | Organisations of 1,722 HR professionals on SHRM's opt-in Voice of Work panel | Fielded 5 to 23 December 2025 | SHRM, The State of AI in HR 2026 |
| intend to launch AI in their HR functions this year | 7% | intend to launch AI in their HR functions this year | Organisations of 1,722 HR professionals on SHRM's opt-in Voice of Work panel | Fielded 5 to 23 December 2025 | SHRM, The State of AI in HR 2026 |
| have not adopted any form of AI in HR and have no plans to do so this year | 54% | have not adopted any form of AI in HR and have no plans to do so this year | Organisations of 1,722 HR professionals on SHRM's opt-in Voice of Work panel | Fielded 5 to 23 December 2025 | SHRM, The State of AI in HR 2026 |
What's stopping it
Eurostat asked the enterprises that considered AI and did not adopt it why not, and published all eight reasons they gave.
The most cited reason EU enterprises gave for considering AI and not adopting it was lack of relevant expertise, at 70.89%. Cost is sixth of the eight, at 33.34%.
- Lack of relevant expertise70.89%
- Lack of clarity about the legal consequences52.52%
- Concerns regarding violation of data protection and privacy48.83%
- Difficulties with availability or quality of the necessary data44.37%
- Incompatibility with existing equipment, software or systems41.39%
- The costs seem too high33.34%
- Ethical considerations23.53%
- AI technologies are not useful for enterprise20.68%
Select any bar for what its figure means.
Eurostat's own wording, in full. Respondents could give more than one reason.
How this was measured
Enterprises that considered AI and did not adopt it, EU
National statistics (tier 1)
2025
| Category | Figure | What it measures | Basis | Year | Source |
|---|---|---|---|---|---|
| Lack of relevant expertise | 70.89% | of enterprises that considered AI and did not adopt gave this reason | Enterprises that considered AI and did not adopt it, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Lack of clarity about the legal consequences | 52.52% | of enterprises that considered AI and did not adopt gave this reason | Enterprises that considered AI and did not adopt it, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Concerns regarding violation of data protection and privacy | 48.83% | of enterprises that considered AI and did not adopt gave this reason | Enterprises that considered AI and did not adopt it, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Difficulties with availability or quality of the necessary data | 44.37% | of enterprises that considered AI and did not adopt gave this reason | Enterprises that considered AI and did not adopt it, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Incompatibility with existing equipment, software or systems | 41.39% | of enterprises that considered AI and did not adopt gave this reason | Enterprises that considered AI and did not adopt it, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| The costs seem too high | 33.34% | of enterprises that considered AI and did not adopt gave this reason | Enterprises that considered AI and did not adopt it, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| Ethical considerations | 23.53% | of enterprises that considered AI and did not adopt gave this reason | Enterprises that considered AI and did not adopt it, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
| AI technologies are not useful for enterprise | 20.68% | of enterprises that considered AI and did not adopt gave this reason | Enterprises that considered AI and did not adopt it, EU | 2025 | Eurostat, Use of artificial intelligence in enterprises |
Who's using it
Away from the business surveys: who uses these tools personally, in which markets, and at what age.
64.0% of Dubai's working-age population used a generative AI product in the second half of 2025. The global average across roughly 150 economies is 16.3%.
Select any market for what its figure means.
- Dubai64.0%
- Singapore60.9%
- London38.9%
- Toronto35.0%
- New York28.3%
- Riyadh26.2%
- Johannesburg21.1%
Select any bar for what its figure means.
How this was measured
Working-age population
Industry survey (tier 3)
H2 2025
Each dot marks a market with a published figure and is sized by it. Dot area is a coarse signal and is not meant to be read as a value; the exact figures are in the ranked chart beside it. A market with no dot is one with no published figure, not one with a figure of zero.
| Category | Figure | What it measures | Basis | Year | Source |
|---|---|---|---|---|---|
| Dubai, United Arab Emirates | 64.0% | of the working-age population used a generative AI product | Working-age population, United Arab Emirates | H2 2025 | Microsoft AI Economy Institute |
| Singapore, Singapore | 60.9% | of the working-age population used a generative AI product | Working-age population, Singapore | H2 2025 | Microsoft AI Economy Institute |
| London, United Kingdom | 38.9% | of the working-age population used a generative AI product | Working-age population, United Kingdom | H2 2025 | Microsoft AI Economy Institute |
| Toronto, Canada | 35.0% | of the working-age population used a generative AI product | Working-age population, Canada | H2 2025 | Microsoft AI Economy Institute |
| New York, United States | 28.3% | of the working-age population used a generative AI product | Working-age population, United States | H2 2025 | Microsoft AI Economy Institute |
| Riyadh, Saudi Arabia | 26.2% | of the working-age population used a generative AI product | Working-age population, Saudi Arabia | H2 2025 | Microsoft AI Economy Institute |
| Johannesburg, South Africa | 21.1% | of the working-age population used a generative AI product | Working-age population, South Africa | H2 2025 | Microsoft AI Economy Institute |
Chatbot use falls with age, from 66% of US adults aged 18 to 29 to 23% of those 65 and over. Pew reports the gender gap on using them at all has closed, while a gap on daily use remains.
18 to 29
30 to 49
50 to 64
65 and over
Select any column for what its figure means.
- Use chatbots daily7.0 pt
- Use chatbots at all3.0 pt
Select any row for what its gap means.
Pew reports that this is a shift from two years earlier, when men were 11 percentage points more likely than women to say they used AI chatbots.
How this was measured
| Category | Figure | What it measures | Basis | Year | Source |
|---|---|---|---|---|---|
| 18 to 29 | 66% | of US adults in this age group use AI chatbots | US adults, n = 5,119 | Surveyed 17 to 23 February 2026 | Pew Research Center, How opinions and use of AI differ by age |
| 30 to 49 | 61% | of US adults in this age group use AI chatbots | US adults, n = 5,119 | Surveyed 17 to 23 February 2026 | Pew Research Center, How opinions and use of AI differ by age |
| 50 to 64 | 42% | of US adults in this age group use AI chatbots | US adults, n = 5,119 | Surveyed 17 to 23 February 2026 | Pew Research Center, How opinions and use of AI differ by age |
| 65 and over | 23% | of US adults in this age group use AI chatbots | US adults, n = 5,119 | Surveyed 17 to 23 February 2026 | Pew Research Center, How opinions and use of AI differ by age |
| Category | Men | Women | Difference | Basis | Year | Source |
|---|---|---|---|---|---|---|
| Use chatbots daily | 27% | 20% | 7.0 percentage points | US adults, n = 5,119 | Surveyed 17 to 23 February 2026 | Pew Research Center, The gender gap in AI |
| Use chatbots at all | 50% | 47% | 3.0 percentage points | US adults, n = 5,119 | Surveyed 17 to 23 February 2026 | Pew Research Center, The gender gap in AI |
What it's doing to work
What the people using AI say it has done to their work, across every industry rather than in one or two of them.
40% of US workers who have used AI chatbots for work say they were highly helpful in doing things more quickly, and 29% say the same about the quality of their work. Among workers who use them at least a few times a month, those figures are 54% and 41%.
40%
said AI chatbots were extremely or very helpful in allowing them to do things more quickly.
29%
said the same about improving the quality of their work.
- Highly helpful in doing things more quickly29.0 pt
- Highly helpful in improving the quality of their work25.0 pt
Select any row for what its gap means.
The reported benefit tracks how much the tool is actually used: both measures are roughly double among workers who use chatbots at least a few times a month compared with those who rarely do. Pew asked employed adults across all industries, so this is not a figure about one sector.
How this was measured
US workers, split by how often they use AI chatbots at work
Industry survey (tier 3)
Surveyed 7 to 13 October 2024
Pew Research Center's American Trends Panel, Wave 157. 5,395 employed US adults across all industries responded, fielded 7 to 13 October 2024. This is a nationally representative panel of US adults working full or part time, so it answers the general workplace question rather than a sector one; it is one country, and it is a survey panel rather than a statistical business register.
| Category | Use chatbots at least a few times a month | Rarely use chatbots | Difference | Basis | Year | Source |
|---|---|---|---|---|---|---|
| Highly helpful in doing things more quickly | 54% | 25% | 29.0 percentage points | US workers, split by how often they use AI chatbots at work | Surveyed 7 to 13 October 2024 | Pew Research Center, Workers' experience with AI chatbots in their jobs |
| Highly helpful in improving the quality of their work | 41% | 16% | 25.0 percentage points | US workers, split by how often they use AI chatbots at work | Surveyed 7 to 13 October 2024 | Pew Research Center, Workers' experience with AI chatbots in their jobs |
| Category | Figure | What it measures | Basis | Year | Source |
|---|---|---|---|---|---|
| said AI chatbots were extremely or very helpful in allowing them to do things more quickly | 40% | said AI chatbots were extremely or very helpful in allowing them to do things more quickly | US workers who have used AI chatbots for work | Surveyed 7 to 13 October 2024 | Pew Research Center, Workers' experience with AI chatbots in their jobs |
| said the same about improving the quality of their work | 29% | said the same about improving the quality of their work | US workers who have used AI chatbots for work | Surveyed 7 to 13 October 2024 | Pew Research Center, Workers' experience with AI chatbots in their jobs |
Across every industry, the people using AI report it helping: 65% of US employees in organisations that have implemented it say the effect on their productivity has been positive, and 65% say the value of their work has risen. Worker confidence in the technology fell 18% in the same period.
Six populations from five publishers, so these are stated rather than charted: no two of them share a denominator. Microsoft and Glean both sell AI software and both surveyed people who already use AI, which is marked on those cards. ManpowerGroup is here because a block built only from vendor research reads as marketing.
65%
say AI has had a positive effect on their productivity and efficiency at work.
65%
said the value of their work has increased since AI tools arrived.
40%
said AI chatbots were extremely or very helpful in allowing them to do things more quickly.
Vendor researchAI users only
66%
say AI has allowed them to spend more time on high-value work.
Vendor researchAI users only
44%
say AI is fairer than their human manager.
18%
decline in overall worker confidence, in the same year regular AI use rose 13% to 45%.
How this was measured
Six different populations, named on every card. None of them share a denominator, which is why none of them share an axis
Industry survey (tier 3)
October 2024 to May 2026
Two of these six come from companies that sell AI software, and both surveyed people who already use it rather than a workforce. Microsoft screened out anyone who never uses AI; Glean states its sample skews higher on AI adoption, seniority and technology than the broader working population. Both are marked on the card. ManpowerGroup attributes its confidence decline to newly introduced questions about AI proficiency, so that figure is not a like-for-like comparison with the previous year.
- OECD, The impact of AI on the workplace: the source of all five measures, and restricted to finance and manufacturing by survey design. The OECD's own 2024 synthesis repeats these figures citing the same 2023 survey.
- Eurofound, European Working Conditions Survey 2024: covers all sectors and 36,644 workers, and measures job quality against exposure to algorithmic management rather than what AI did to the work.
- Stanford HAI, AI Index: reports effects by business function for marketing and sales, service operations, supply chain and software engineering, and publishes none for finance or human resources.
- Performance, the value of the work, and fairness in management do have all-sector coverage from named publishers. Those figures are in the block above, on their own populations.
- Mental health and wellbeing: no all-sector equivalent found. Eurofound is the nearest and asks a different question.
- Physical health and safety: no all-sector equivalent was looked for. The measure is manufacturing-shaped by nature, and substituting something else for it would be worse than leaving it sector-bound.
| Category | Figure | What it measures | Basis | Year | Source |
|---|---|---|---|---|---|
| say AI has had a positive effect on their productivity and efficiency at work | 65% | say AI has had a positive effect on their productivity and efficiency at work | US employees in organisations that have implemented AI | As of May 2026 | Gallup, Global Indicator: Artificial Intelligence |
| said the value of their work has increased since AI tools arrived | 65% | said the value of their work has increased since AI tools arrived | 5,875 US workers across all sectors | Surveyed March and April 2026 | SHRM, Navigating AI in the Workplace 2026 |
| said AI chatbots were extremely or very helpful in allowing them to do things more quickly | 40% | said AI chatbots were extremely or very helpful in allowing them to do things more quickly | US workers who have used AI chatbots for work | Surveyed 7 to 13 October 2024 | Pew Research Center, Workers' experience with AI chatbots in their jobs |
| say AI has allowed them to spend more time on high-value work | 66% | say AI has allowed them to spend more time on high-value work | 20,000 knowledge workers who already use AI, 10 markets | Fielded 18 February to 20 April 2026 | Microsoft, 2026 Work Trend Index |
| say AI is fairer than their human manager | 44% | say AI is fairer than their human manager | 6,000 full-time digital workers, US, UK and Australia | Surveyed December 2025 to January 2026 | Glean, Work AI Index 2026 |
| decline in overall worker confidence, in the same year regular AI use rose 13% to 45% | 18% | decline in overall worker confidence, in the same year regular AI use rose 13% to 45% | 13,918 workers across 19 countries | Fieldwork 1 September to 1 October 2025 | ManpowerGroup, Global Talent Barometer 2026 |
AI use at work is not confined to a couple of industries, but it is very uneven across them: 76% of US employees in technology or information systems used AI in their role, against 33% in retail.
- Technology or information systems76%
- Finance58%
- Professional services57%
- Manufacturing38%
- Healthcare37%
- Retail33%
Select any bar for what its figure means.
Gallup asks every industry the same question, which is what makes the distance between these six readable. It measures whether someone used AI in their role at all, not how much.
How this was measured
US employees, by industry
Industry survey (tier 3)
Q3 2025, surveyed 5 to 19 August
Gallup Workforce results, based on a nationally representative survey of 23,068 US adults employed full and part time, conducted by web on the Gallup Panel between 5 and 19 August. Gallup reports 45% of employees using AI at least a few times a year in Q3 2025, and daily use at around 10% of the American workforce.
| Category | Figure | What it measures | Basis | Year | Source |
|---|---|---|---|---|---|
| Technology or information systems | 76% | of employees in this industry used AI in their role a few times a year or more | US employees, by industry | Q3 2025, surveyed 5 to 19 August | Gallup, AI Use at Work Rises |
| Finance | 58% | of employees in this industry used AI in their role a few times a year or more | US employees, by industry | Q3 2025, surveyed 5 to 19 August | Gallup, AI Use at Work Rises |
| Professional services | 57% | of employees in this industry used AI in their role a few times a year or more | US employees, by industry | Q3 2025, surveyed 5 to 19 August | Gallup, AI Use at Work Rises |
| Manufacturing | 38% | of employees in this industry used AI in their role a few times a year or more | US employees, by industry | Q3 2025, surveyed 5 to 19 August | Gallup, AI Use at Work Rises |
| Healthcare | 37% | of employees in this industry used AI in their role a few times a year or more | US employees, by industry | Q3 2025, surveyed 5 to 19 August | Gallup, AI Use at Work Rises |
| Retail | 33% | of employees in this industry used AI in their role a few times a year or more | US employees, by industry | Q3 2025, surveyed 5 to 19 August | Gallup, AI Use at Work Rises |
Start with
the diagnostic.
A scoping conversation costs nothing and ends with a straight answer about whether there is enough here to be worth doing. If there is not, we will tell you.