AI in the Workplace Statistics: Usage, Shadow AI and Trust

Last Updated: October 5, 2026

About half of U.S. employees now use AI at work, but only 15% use it daily and just 25% say their employer has shared a clear AI plan. Meanwhile, 57% of workers worldwide admit hiding their AI use, and IBM found shadow AI in 43% of studied data breaches in 2026, up from 20%. Latest data as of October 2026.

The most useful number in AI in the workplace statistics is not how many people use AI. It is how many use it without telling anyone.

Adoption has crossed the halfway mark in the U.S., yet policies, training, and trust have not kept pace. This page collects the numbers that matter, each with its population, survey date, and source. Where surveys disagree, the sections below explain why instead of picking a winner.

💡 Not sure which AI tool is actually right for your business?

Get the free guide, Which AI Tool Should You Actually Use? — a straightforward breakdown of the leading AI tools to help you pick the right one for your needs.

Subscribe to AI Business Weekly for the guide, plus daily coverage of AI trends, acquisitions, and product launches.

Get the Free Guide →

About half of U.S. employees use AI at work at least a few times a year, based on Gallup's May 2026 data. Only 15% use it daily, and 30% use it a few times a week or more. Measured by past-week use, the St. Louis Fed puts the figure at 39.2%.

Every survey measures something slightly different, which is why the headline numbers range from 21% to 75%.

The spread is mostly definition and population, not disagreement. Microsoft and BCG surveyed white-collar and knowledge workers, who skew toward desk jobs where AI tools fit easily, while Pew's 21% covers all U.S. workers and a narrower question, and was fielded earlier. Treat roughly 50% as "has used AI at work" and 30% as "uses it habitually."

For the business-wide picture, see our AI adoption statistics page.

Technology workers lead, with 42% using AI daily in Gallup's May 2026 data, versus 27% in finance and 22% in professional services. Every other industry sits between 9% and 15%. Employers in the same three industries were also the most likely to encourage chatbot use, per Pew.

The ranking is identical in both surveys, even though they measure different things and were fielded about 19 months apart. Gallup attributes role differences to how well mainstream AI tools match the tasks people actually do, and desk-based work matches best.

What do workers actually do with it? Among U.S. workers who use AI chatbots, 57% research or find information, 52% edit written content, 47% draft reports or documents, and 40% summarize information (Pew, survey fielded October 2024, published February 2025).

Yes, though the gap is narrowing. Gallup's May 2026 data shows 51% of leaders use AI frequently, versus 36% of managers and 26% of individual contributors. BCG found 74% of frontline employees are regular AI users in 2026, up from 51% a year earlier.

The gap runs in both directions. In McKinsey's survey of 3,613 employees and 238 C-suite executives (October to November 2024, 81% U.S. respondents), 13% of employees said they use generative AI for at least 30% of their daily work, while executives estimated that figure at 4% (McKinsey, Superagency in the Workplace). Leaders use AI more than their staff, yet underestimate how much their staff use it, so policy gets written for a workforce that does not exist.

One caution on the numbers: Gallup's February 2026 survey, limited to employees whose organizations have AI tools available, found 67% of leaders and 46% of individual contributors using AI daily or a few times a week. That base is narrower than the May figures above, which likely explains the higher percentages.

Shadow AI means employees using AI tools their employer has not approved. It is widespread: Microsoft and LinkedIn found 78% of AI users brought their own tools to work in 2024, and KPMG's global study found 57% of employees hide their AI use. About one in three reaches AI through personal accounts, per Cyberhaven.

Cyberhaven also tracked whether employees signed in with personal or company accounts, a basic form of shadow AI detection.

These figures come from a security vendor's customer base, so they reflect companies that already monitor data flows, and the study summary does not state its collection period. Read them as directional. The pattern is still clear: when official tools do not exist or do not fit, shadow AI is a provisioning problem before it is a discipline problem.

IBM's 2026 Cost of a Data Breach study found shadow AI in 43% of studied breaches, more than double the 20% in its 2025 study. More than two-thirds of organizations said they had no governance processes in place to limit shadow AI.

The cost attaches to missing governance, not to AI use itself. The 2025 and 2026 studies sampled different organizations, so treat the jump from 20% to 43% as a study-level signal, not a precise one-year trend. For the practical side, see our guide to the risks of using AI at work, our AI cybersecurity statistics, and the AI governance framework breakdown.

Trust is split. Pew found 52% of U.S. workers worried about future AI use in the workplace and 36% hopeful (October 2024). McKinsey found 71% of employees trust their employer to deploy AI safely, and KPMG found only 46% of people worldwide are willing to trust AI systems.

The surveys cover different populations, but together they suggest workers trust their employer more than they trust AI itself. That helps explain concealment: if an employee believes AI use signals replaceability, hiding it is a rational reaction to a leadership gap rather than a character flaw. Job-loss estimates sit outside this page, so see our AI job market statistics for displacement data.

Most do not, or workers do not know about them. KPMG's global study found 40% of employees have workplace policy or guidance on generative AI and 47% have received AI training. Gallup found 47% of U.S. employees say their organization has integrated AI, but only 25% say it has communicated a clear plan.

Training also correlates with use. BCG's 2025 survey found regular AI use notably higher among employees who received at least five hours of training and had access to in-person coaching. Combined with its finding that more than half of employees will work around missing tools, that makes the rollout step hard to skip, and our guide on how to implement AI in business covers the policy and training steps in order.

About 52% of U.S. employees use AI at work at least a few times a year, and 15% use it daily, according to Gallup's May 2026 data. Pew found 38% of employed adults use chatbots for work tasks (February 2026). Definitions drive the gap, as the comparison table above shows. For company-level figures, see AI adoption statistics.

Shadow AI is the use of AI tools at work without employer approval or oversight, such as employees pasting company data into personal ChatGPT, Claude, or Perplexity accounts. Microsoft and LinkedIn found 78% of AI users brought their own tools to work in 2024. The term mirrors "shadow IT," which describes unapproved software.

Shadow AI increases data exposure and breach cost. IBM's 2026 study found shadow AI in 43% of studied breaches, up from 20% in its 2025 study, and the 2025 study tied it to as much as $670,000 in added cost. Governance gaps, not AI itself, drove the cost. See our AI governance framework guide.

Among U.S. workers who use AI chatbots, 57% use them to research or find information, 52% to edit written content, 47% to draft reports or documents, and 40% to summarize information (Pew, October 2024 survey). Writing and research tasks lead, which explains why desk-based roles adopt AI fastest.

Gallup found 65% of U.S. employees say AI has had a positive effect on their productivity, with 17% calling it extremely positive (May 2026). Yet only 14% strongly agree AI has transformed how work gets done. For measured output and time-saving data, see our AI productivity statistics.

Technology leads, with 42% of employees using AI daily, followed by finance at 27% and professional services at 22% (Gallup, May 2026). All other industries range from 9% to 15%. Pew's 2024 survey shows the same ranking for employer encouragement of chatbot use.

Trust is mixed. Pew found 52% of U.S. workers worried about future workplace AI use and 36% hopeful (October 2024), while McKinsey found 71% of employees trust their employer to deploy AI safely (late 2024). Worry centers on AI's future impact, not on employer intent.

AI in the workplace statistics for 2026 point to a gap between use and management. About half of U.S. employees use AI at work, but only 15% use it daily, and only 25% say their employer has communicated a clear plan. Leaders use AI more than staff yet underestimate how much staff use it.

The result is shadow AI: 57% of workers worldwide hide their AI use, 78% of AI users brought their own tools in Microsoft's 2024 survey, and IBM found shadow AI in 43% of studied breaches in 2026, up from 20% a year earlier. Training, clear policies, and manager support are the levers the data shows moving usage and trust.

By Sameer Khan

AI-assisted. Researched, reviewed, and edited by Sameer Khan before publication.