AI in the Workplace Statistics (2026)

Fifty-two percent of US employees used AI in their role at least a few times a year in May 2026, 30% used it a few times a week or more, and 15% used it daily. Three years earlier the figures were 21%, 11% and 4%. On the employer side, 19.8% of US businesses reported using AI in any business function in the two weeks to 3 May 2026, rising to 37% of firms with 250 or more employees. Nine in ten executives surveyed say they have seen no measurable effect on their own firm's productivity or employment yet.

The pages that rank for this question quote a 2024 figure of 75% as current, mix Gallup's quarterly readings without dates, present a six-country vendor survey as US data and treat a 2023 forecast of displaced jobs as a measurement. This page uses Gallup's quarterly panel with each quarter dated, Pew's worker surveys, the Census Bureau's business survey, the St. Louis Fed's adoption tracker, the five field experiments on productivity and the surveys on policy and monitoring, with sample and date on every figure.

How many workers use AI, and how often

Gallup has asked the same question of its US employee panel every quarter since 2023, which makes it the one series with a trend. The second-quarter 2026 reading is from 22,573 employees surveyed in May.

Use is uneven. In the fourth quarter of 2025, 69% of leaders, 55% of managers and 40% of individual contributors used AI at all; 66% of workers in remote-capable roles against 32% in roles that cannot be done remotely; and 77% in technology, 64% in finance and 33% in retail. Daily use in technology reached 42% by May 2026. Forty-three percent of US employees never use AI at work.

Two other measures bracket Gallup's. Pew Research Center's September 2025 survey of 5,010 workers found 21% saying at least some of their work is done with AI, up from 16% a year earlier, with 65% not using it much or at all. The question is stricter than Gallup's, which is why the figure is lower. The St. Louis Fed's Real-Time Population Survey found 45.2% of employed adults aged 18 to 64 had used generative AI for work in April 2026, up from 33.3% in July 2024, and 61.8% had used it at all. Its researchers calculated that generative AI reached 54.6% adoption three years after launch, against 19.7% for the personal computer and 30.1% for the internet at the same point.

How many businesses use AI

The Census Bureau's Business Trends and Outlook Survey, a biweekly probability sample of about 1.2 million employer businesses, found 19.8% using AI in any business function in the two weeks ending 3 May 2026, with 20% to 23% expecting to within six months. By size, 37% of firms with 250 or more employees, 32% of those with 100 to 249 and under 20% of those with four or fewer. By sector, 39.7% in information, 33.9% in finance and insurance and about 14% in retail. The question changed in November 2025 from use "in producing goods or services" to use "in any business function," so the earlier readings of 4% to 6% are not comparable with the 2026 series.

The New York Fed's 2025 Small Business Credit Survey of 5,248 employer firms found 46% where the owner or employees use AI tools, 77% of users reporting no change in labor costs and 31% reporting higher sales. AI-using small firms were more likely to expect to add staff. The global figure that circulates, 88% of organizations using AI in at least one function, is from McKinsey's self-selected survey of about 2,000 respondents and is not comparable with the Census probability sample. Gallup's employee-side measure found 47% of US workers saying their organization has integrated AI tools in May 2026, up from 38% in late 2025, with about one in five unsure.

What workers use it for

Among US employees who use AI, Gallup's May 2026 data show 51% using it for writing and editing, 49% for search or research, 39% for general problem-solving, 18% for data analytics, 17% for slide decks, 16% for coding and 16% for process automation. Pew's chatbot users most often used them for research at 57%, editing at 52% and drafting at 47%. Anthropic's analysis of a million Claude conversations in August 2025 found 39% were full task delegation, up from 27% eight months earlier, and that coding was 36% of use; Microsoft's analysis of 105,000 Copilot conversations in February 2026 found 49% supporting cognitive work and 19% working with people. Platform data describe the users of one product and should be read that way.

What the field experiments found

Five studies have measured output rather than asking about it, and they do not agree.

• Customer support, 5,179 agents. Brynjolfsson, Li and Raymond's staggered rollout of a generative AI assistant raised issues resolved per hour by 14% on average, by 34% for novice agents and barely at all for experienced ones. The data are from 2020 and 2021 and the paper appeared in the Quarterly Journal of Economics in 2025.
• Professional writing, 453 people. Noy and Zhang's randomized trial in Science found ChatGPT cut time on writing tasks by 40% and raised quality 18%, narrowing the gap between weaker and stronger writers.
• Consultants, 758 people. Dell'Acqua and colleagues' experiment at BCG found GPT-4 users completed 12.2% more tasks 25.1% faster at 40% higher quality on tasks inside the model's capability, and were 19 to 23 points less likely to get the answer right on a task outside it.
• Experienced developers, 16 people. METR's 2025 trial found developers 19% slower with early-2025 tools on real tasks in their own repositories, while believing they were 20% faster. METR says the result is now out of date and its 2026 follow-up found a smaller, uncertain effect.
• Executives, about 6,000. Yotzov, Barrero, Bloom, Davis and colleagues' 2026 survey in the US, UK, Germany and Australia found 69% of firms actively using AI and about nine in ten executives reporting no measurable effect on their own firm's productivity or employment over three years, with expectations of 1.4% higher productivity and 0.7% lower employment over the next three.

Self-reports sit between those. The St. Louis Fed's pooled 2025 data put time saved at 5.4% of work hours for users, about 2.2 hours in a 40-hour week, and 1.6% across all workers including non-users, implying an aggregate productivity gain of up to 1.3% since late 2022. In Gallup's data, 65% of employees at AI-adopting organizations say AI has improved productivity, rising from 45% among those using it for one or two purposes to 90% among those using it for seven or more. Microsoft's 2026 survey of 20,000 knowledge workers who already use AI, in ten countries, found 66% saying it frees time for higher-value work. Our remote work productivity statistics page covers the parallel measurement problem.

Worker attitudes and job worry

Pew's October 2024 survey of 5,273 workers found 52% worried about the future use of AI in the workplace, 36% hopeful and 33% overwhelmed; 32% expected fewer job opportunities for themselves long term and 6% more. Yet the share who think AI or automation is likely to eliminate their own job within five years has barely moved in Gallup's data: 15% in 2023, 2024 and 2025 and 18% in early 2026, rising to 23% at AI-adopting organizations. The general public is more pessimistic than workers about themselves: Pew's June 2026 survey of 3,488 adults found 71% expecting AI to reduce the number of US jobs over 20 years, up from 64% in 2024, and 52% more concerned than excited about AI in daily life, up from 37% in 2021. AI experts disagree with the public: 73% expect AI to improve how people do their jobs against 23% of the public.

Among non-users in Gallup's early 2026 data, the reasons were preferring their current way of working at 46%, data privacy and compliance concerns at 43%, ethical opposition at 43% and not believing AI can help their work at 39%. Our Gen Z workplace statistics page covers the youngest workers, among whom 48% say AI's risks outweigh its benefits.

AI and engagement

Gallup's July 2026 analysis of 43,262 US employees found engagement 6 points higher at AI-adopting organizations, 15 points higher where the organization had communicated a clear AI plan, and 48% where the manager actively supports AI use against 30% where the manager does not. Weekly users were 8 points more engaged than non-users. In AI-implementing workplaces, 25% of employees said culture had worsened over the past year, 24% that it had improved and 51% that it was unchanged. Our employee engagement statistics page sets those against the flat national rate of 31%.

Training, policy and strategy gaps

Twenty-five percent of US employees strongly agree their organization has communicated a clear AI plan or strategy, up from 22% in 2025, per Gallup's AI indicator. In 2025, 44% said their organization had begun integrating AI while 30% said it had guidelines or formal policies, a 14-point gap. Only 36% of employees at AI-integrating organizations strongly agree their manager supports the team's use, and those who do are 1.7 times as likely to use it weekly. Where employees strongly agree policies are clear, frequent use is 68% against 47%.

Employers report more policy than employees see. Littler's 2026 survey of more than 300 US executives and in-house counsel found 68% with a formal AI policy, up from 38% a year earlier, 55% with a formal review process for AI tools and 54% restricting what information can be entered. EisnerAmper's 2025 survey of 1,017 degree-holding desk workers who use AI found 36% saying their company has a formal policy. Among 102 chief human resources officers Gallup surveyed in early 2026, 57% were providing AI training for managers and 50% were not confident in managers' ability to guide AI use. Our employee training statistics page covers how little of training spend goes to AI.

Employer monitoring of AI use

In the EisnerAmper survey, 22% of AI-using desk workers said their employer actively monitors their AI use, 41% inform their manager or seek permission before using it, 60% rely on free public tools and 28% would use AI even if it were banned. Eighty-four percent of managers acknowledged some AI use on their teams. The sample is unweighted and limited to degree holders at desks, so it describes that population only. Among Gallup's non-users, 43% cite data privacy, security and compliance as a reason. Our employee monitoring statistics page covers the wider monitoring picture.

Figures to stop using

• "75% of knowledge workers use AI at work." Microsoft's May 2024 survey of 31 countries. The US all-worker figure is 52% by Gallup's broad measure or 21% by Pew's strict one.
• Gallup figures without quarters. 40%, 45%, 46%, 50% and 52% are five different quarters of one question.
• "Pew says 16% but Gallup says 40%." Different questions and dates. Pew's own update is 21%.
• "92 million jobs will be displaced." The World Economic Forum's January 2025 employer-expectation survey, which also projects 170 million created, for a net 78 million by 2030.
• "Goldman Sachs: 300 million jobs lost." The 2023 note said 300 million jobs globally were exposed to automation, not lost.
• "Only 4% to 6% of businesses use AI." Census readings under the old question wording. The May 2026 figure is 19.8%.
• "AI makes developers 19% slower." Sixteen developers with early-2025 tools; METR says it is out of date.
• "88% of companies use AI" presented as US. A self-selected global survey against a probability sample at 19.8%.

Key takeaways

• 52% of US employees used AI at work at least a few times a year in May 2026, 30% weekly or more and 15% daily, up from 21%, 11% and 4% in 2023.
• 19.8% of US businesses used AI in any function in May 2026, rising to 37% of firms with 250 or more employees. The 88% figure is a global self-selected survey.
• Use is uneven: 69% of leaders against 40% of individual contributors, 66% in remote-capable roles against 32%, 77% in technology against 33% in retail.
• Field experiments found gains of 14% in customer support, 40% less time on writing and 25% faster consulting work, alongside a 19% slowdown for experienced developers and nine in ten executives reporting no measurable firm-level effect.
• Users report saving 5.4% of work hours, about 2.2 hours a week; across all workers it is 1.6%.
• 52% of workers are worried about AI at work but only 18% think it is likely to eliminate their own job within five years. 71% of the public expects fewer US jobs over 20 years.
• Engagement is 48% where managers support AI use against 30% where they do not, but only 25% of employees say their employer has a clear AI plan.
• 68% of employers say they have a formal AI policy; 36% of AI-using desk workers say their company has one, and 22% say their use is monitored.