AI at Work

Half of American Workers Now Use AI, but Most Gains Are Small

By Lena Ortiz / November 5, 2026 / Updated November 8, 2026

Smiling woman seated at a table with her laptop, looking at the camera

Quick answer: Yes, about half of U.S. workers now use AI in their jobs, but “use” is a generous word. Gallup found 50% in February 2026 and 52% in May 2026 using AI at least a few times a year, while only about 15% used it daily. Most people who report a benefit describe faster work on individual tasks, and the broader evidence points to modest gains on average so far. Think “handy helper,” not “workplace revolution.”

Key takeaways

Gallup’s May 2026 survey of 22,573 employed U.S. adults found 52% use AI in their role at least a few times a year, 30% a few times a week or more, and 15% daily.
“Half use AI” counts anyone who uses it a few times a year. Heavy daily use is a much smaller club.
Among AI users in a separate Gallup-involved study, 63% said AI helps them work faster. These are self-reported perceptions, not measured output.
Executives in an NBER working paper estimated AI added only about 0.2% to annual labor productivity growth over the past three years.
Gains are uneven across jobs, industries and education levels.

What “half of workers use AI” actually measures

The headline comes from Gallup’s quarterly workforce survey. The first-quarter 2026 wave, published April 12, 2026, surveyed 23,717 employed U.S. adults between February 4 and 19, 2026. Half said they use AI in their role at least a few times a year, up from 46% the previous quarter. The second wave, fielded May 6 to 20 with 22,573 respondents, put the figure at 52%. Both carry a margin of error of about plus or minus 0.9 percentage points.

Respondents come from the probability-recruited Gallup Panel, weighted to match U.S. demographics. One thing we could not find on the pages we read is the exact wording of the AI-use question, so we cannot say precisely what counted as “using AI.” That matters, because someone who asked a chatbot to reword an email once last month and someone who codes with an assistant all day both land in the “use” bucket.

Gallup measure (U.S. employees)	Q1 2026 (fielded February 4-19)	Q2 2026 (fielded May 6-20)
Use AI at least a few times a year	About 50%	52%
Use AI a few times a week or more	28%	30%
Use AI daily	13%	15%
Say their organization has integrated AI tools	41%	47%

As of November 2026, Q2 is the latest Gallup wave we found; we did not find a third-quarter release.

How often people use it

The gap between “ever” and “daily” is where the story lives. On Gallup’s indicator page, daily use ranges widely by industry: 42% in technology, 27% in finance and 22% in professional services, with the other seven industries listed between 9% and 15%.

A separate study, the American Job Quality Study from Jobs for the Future, the Families and Workers Fund and Gallup, shows a similar picture. In a survey fielded January 26 to March 24, 2026 (15,482 employees), 28% used AI at least weekly and 54% said they did not use it in their role. Regular use was more common among managers (37%, versus 25% of individual contributors) and college graduates (40%, versus 17% of those without a degree). In healthcare support, manufacturing production and retail sales, 70% to more than 80% never use AI at work.

What people say they gain

Most users say it helps. In the first-quarter data, 65% of employees at organizations that have adopted AI said it improved their productivity and efficiency, 16% called the effect extremely positive, and fewer than one in 10 said it was negative.

The American Job Quality Study asked 7,845 AI users what got better. 63% said they work faster, 56% said they find more creative solutions, 47% said they have more time for the most interesting parts of the job, and 46% said they produce higher-quality work. Regular users reported bigger benefits than occasional ones, such as 79% versus 38% saying AI makes them faster. About 31% said their employer had asked them to take on more responsibilities because of their AI use, which is the “congratulations, here is more work” effect.

In Gallup’s May wave, the top uses among AI users were writing and editing (51%) and search or research (49%). People who use AI for more purposes rate it higher, but Gallup warns that does not prove more uses cause bigger gains. People who see value may simply try more things.

Why most gains look small

Feeling faster is not the same as the company getting more done. Gallup’s first-quarter data shows the gap: only about one in 10 employees at AI-adopting organizations strongly agreed AI had transformed how work gets done where they work. Benefits are concentrated in individual tasks. Other research gives a cooler view, so watch the dates and caveats.

Evidence	What it found	Caveat
NBER Working Paper 34836, “Firm Data on AI” (revised October 2026)	In a survey of nearly 9,000 senior executives in seven countries, 78% of firms use AI. Executives reported essentially no net effect on employment over three years and a boost to annual labor productivity growth of about 0.2%, with 0.6% per year expected over the next three years.	Executive estimates, not measured output. A working paper, meaning research shared before formal peer review.
St. Louis Fed, “The Impact of Generative AI on Work Productivity” (February 27, 2025)	Workers who used generative AI in the prior week saved an average of 5.4% of work hours; counting nonusers, 1.4% of total hours. The authors roughly estimated a 1.1% aggregate productivity gain.	Older data (2024 surveys), self-reported, and saved time may not become extra output.

Put together, adoption is broad, perceived benefits are widespread, and measured economy-wide gains are modest so far. That does not mean they will stay modest, only that dramatic claims have not yet shown up in the numbers we checked.

How to read the next “X% use AI” headline

Who was surveyed: all workers, executives, or only people at companies already using AI?
How “use” was defined: a few times a year, weekly or daily?
When it was fielded. Early 2025 numbers may be stale.
Whether the result is self-reported or measured. “I feel faster” and “we shipped more” are different claims.
The sample size, the margin of error, and whether your subgroup is large enough to say much.
Whether the source admits what it cannot prove. Good ones do.

Frequently asked questions

Do half of American workers really use AI every day?

No. The “half” figure means at least a few times a year. In Gallup’s May 2026 survey, 30% used it a few times a week or more and 15% daily.

Does AI make people more productive?

Many users say so: 65% of employees at AI-adopting organizations in Gallup’s first-quarter data reported better productivity and efficiency. Measured effects look smaller: the St. Louis Fed found recent users saved 5.4% of work hours in 2024 data, and NBER executives estimated a small lift so far.

Who benefits most?

In the American Job Quality Study, regular users reported more benefits than occasional ones, and users with a four-year degree were likelier than those with a high school education or less to say AI makes them faster (68% versus 51%). It cannot prove AI use causes better outcomes.

Is my job at risk because of these numbers?

These surveys are not a forecast for anyone’s job. The NBER executives reported essentially no net employment effect so far. For career questions, talk to a career advisor or your employer.

Before you decide

If you are wondering whether to lean on AI at work, the evidence supports starting small. Pick one repeatable job, such as a first draft, a summary or tidying a spreadsheet, let the tool do a first pass, and check the result before it goes out under your name. Track whether you actually save time over a few weeks, not whether it feels fast.

Follow your employer’s rules about what data can go into an AI tool, and keep your own judgment in the loop. Survey numbers keep moving, so look for the next release before relying on today’s.