More Than Half of US Employees Now Use AI at Work, Gallup Finds

More than half of U.S. employees now use artificial intelligence on the job, according to Gallup, the first time its long-running tracker has crossed the halfway point. Fifty-two percent of workers say they use AI at least a few times a year, up from 27% two years ago, and 15% now use it every day.

The jump is real. A quieter number appears right beside it in the same survey: one in five employees still can’t say whether their own employer has adopted AI at all, and that confusion concentrates hardest among the workers furthest from the executive suite.

AI Use at Work Finally Clears 50%

The survey from Gallup, the polling and analytics firm, released Tuesday, puts workplace AI use at 52%, up from 27% two years ago, the first time more than half of U.S. employees have reported using the technology on the job. A third of employees now use AI a few times a week or more, and 15% use it daily.

Among employees who use the tools, the most common tasks skew toward everyday knowledge work rather than anything exotic:

• Writing and editing – 51% use AI to draft or revise text
• Search or research – 49% use it to find information
• General assistance or problem solving – 39% use it this way

The growth has been steady rather than sudden, and it shows clearly in Gallup’s own quarterly tracking of the same questions:

Adoption isn’t even across industries. Growth in the technology sector, where AI use has run highest, shows signs of leveling off after an explosive rise between 2024 and 2025, while majorities in professional services, higher education and K-12 education now report using AI at least occasionally.

Employees at companies that have already adopted AI were also nearly twice as likely to report large-scale disruption at their workplace over the past year, 27% compared with 17% at organizations that haven’t adopted the tools.

Coding and Automation Score the Highest Marks

Not all AI use pays off equally. Employees who use AI for coding assistance, process automation and building decks or presentations rate its impact on their work the most highly, with more than three-fourths calling the effect somewhat or extremely positive.

Frequency matters more than any single task, though:

• 45% of employees using AI for one or more tasks report a positive hit to their own productivity
• 90% of employees using it for seven or more tasks report the same
• 75%+ rate coding, automation and deck-building specifically as somewhat or extremely positive

The real-world version of that gap shows up at companies already deep into daily use. At Bank of America, 28-year-old investment banker Andrea Tanzi said he uses AI tools every day to synthesize documents and data sets that would otherwise take him hours to review, including the bank’s internal chatbot, Erica, for administrative tasks.

Economists are starting to track the same divide with firm-level data instead of survey answers. A National Bureau of Economic Research working paper published in February is examining firm-level data on AI adoption patterns, cited alongside Gallup’s own research.

Do Employees Even Know Their Company Is Using AI?

Not always. Forty-seven percent of employees say their organization has integrated AI tools to improve productivity or quality, up from 41% the prior quarter. But 20% say they simply don’t know, a share that has barely moved from the 23% Gallup recorded in the third quarter of last year.

That uncertainty isn’t spread evenly across the org chart. Gallup’s role-by-role data shows individual contributors are both the least frequent users of AI and the most likely to be in the dark about whether their company adopted it at all.

Leaders lead in official adoption, and a separate survey found the same holds true in unsanctioned use, with executives outpacing junior staff on unsanctioned AI use. The people writing AI policy are often its heaviest users outside that policy, too.

Shadow AI Is Outrunning Company Policy

The industry has a name for this: “shadow AI,” tools employees use for work without their IT department’s knowledge or approval.

Between 40% and 65% of enterprise employees use AI tools their IT department hasn’t approved, based on data drawn from IBM’s 2025 Cost of a Data Breach Report and cybersecurity firm Netskope’s Cloud and Threat Report 2026.

Shadow AI detections rose fourfold in a single year, according to Verizon’s 2026 Data Breach Investigations Report, and two-thirds of office professionals said they’ve used an AI tool at work despite believing it broke company policy.

Retool, a software development company, surveyed 307 senior technology and security leaders this year and found a widening gap between visibility and governance inside their own organizations. The report describes this year as the point where shadow IT, unauthorized software employees pick up on their own, becomes shadow AI.

Enterprise infrastructure hasn’t caught up either. A separate survey found most firms lack the infrastructure agentic AI requires, which helps explain why tools employees adopt on their own keep outpacing what companies can actually track or secure.

Sitting Out Has Its Own Cost

Gallup data published this month found that workers who don’t use AI are showing up disproportionately among those laid off. About 21% of U.S. employees reported layoffs at their company in the first quarter of 2026, a share that has held roughly steady after nearly tripling between the second quarter of 2022 and the third quarter of 2025.

Sixty-two percent of laid-off workers were not using AI, Gallup found, though only 1% of laid-off employees pointed to AI or automation as the reason they lost their job. Most cited organizational restructuring, role elimination or other downsizing instead.

Confidence in Regulators Keeps Falling

Usage is outrunning trust. Pew Research Center, a nonpartisan polling and research organization, found in a February survey that two-thirds of Americans, 67%, have little or no confidence in the government’s ability to regulate AI effectively, up from 62% in 2024. About six in ten adults said the same about the companies building the technology.

The same poll found 31% of respondents expect AI to have a negative effect on them personally, and 40% expect a negative effect on society, over the next 20 years. Adults under 30 were more likely than older respondents to expect harm.

The shift isn’t uniform across party lines, either. Democrats who lack confidence in government AI regulation rose from 54% in 2024 to 74% this year, while Republicans fell from 70% to 61% over the same span, according to Pew, essentially trading places.

Spending on AI governance is racing to keep up regardless. Gartner forecasts the market for AI governance tools will reach $492 million this year and top $1 billion by 2030, roughly doubling in four years, as companies try to build oversight for tools their own employees already adopted first.