Table of contents
1 in 4 Workers Secretly Avoid AI at Work (2026 Survey)
Report Highlights
- About 1 in 4 workers (26%) have secretly avoided or refused AI at work without telling their manager or team.
- Pride is the top reason workers limit AI (39%), ahead of distrust in its accuracy (33%).
- Nearly half (45%) have avoided an AI tool and done the task by hand instead.
- Just 4% completely trust AI to handle important work tasks, while 39% trust it for small things only.
- Gen Z is the most skeptical generation, with 48% distrusting AI's accuracy, compared with 29% of Gen X.
Most Americans can use AI on the job, but some quietly opt out. In our 2026 survey of 918 employed adults, about 1 in 4 (26%) admitted to avoiding or refusing AI without telling their manager or team.
The reason is rarely fear: workers most often skip AI out of pride in their own work, and the leaders who push it hardest are among the likeliest to dodge it in secret.
To start, we asked respondents whether their companies provide or sanction the use of AI tools.
Nearly two-thirds of workers (64%) said they have access to employer-backed AI at work, whether it's approved for optional use or officially expected. The most common setup, reported by 42%, is tools that are approved but not required. Another 22% have official tools they are expected to use.
A further 28% could use AI but lack formal approval, and just 5% are entirely blocked from it.
Senior leaders are most likely to face required tools, at 36%, versus 19% of individual contributors. Owners/partners lean the other way: 38% could use AI if they wanted, compared with 16% of senior leaders.
Working manually is the most common form of quiet AI pushback, reported by 45% of workers.
Still, 42% report none of the listed behaviors.
Smaller groups have delayed AI as long as possible (17%) or talked others out of using it (14%). A few have gone further, claiming a tool "didn't work" to avoid it (5%) or saying they used AI when they hadn’t (4%).
Senior leaders push back hardest: 56% have done a task by hand rather than using AI, compared with 38% of owners/partners. Gen Z (25%) and millennials (21%), on the other hand, have delayed AI as long as they could, compared with 9% of baby boomers. Gen Z and millennials are also the most likely to have talked others out of using AI, at 19%, compared with 8% of Gen X.
We also asked workers, in their own words, about the last time they chose to skip AI and why.
-
The most common theme is the protection of a personal voice. Many want emails and messages to sound genuine rather than generic, especially for sensitive or personal notes.
-
Others skip AI when a task is too simple to be worth the effort of prompting, reviewing, and correcting the result.
-
Accuracy and trust concerns surface often, with workers citing past errors or hallucinations in specialized work.
-
Privacy is another driver, as some refuse to enter financial, medical, or confidential client data into AI tools.
A final theme is skill preservation, with workers keeping their own abilities sharp rather than outsourcing core work.
The leading brake on AI is professional pride. Pride in doing the work themselves tops the list at 39%, ahead of the 33% who distrust AI's accuracy.
Other prominent brakes include time spent on fixing output (29%), ethical concerns (22%), privacy concerns (21%), and energy and water use (21%).
Interestingly, Gen Z is the most skeptical generation across nearly every reason. They lead in distrust of accuracy (48% vs 29% of Gen X), ethical concerns (34% vs 14% of baby boomers), and energy and water use (33% vs 12% of baby boomers).
Women cite ethical concerns (26% vs 18%) and energy and water use (26% vs 15%) more than men, while men more often call AI's output quality poor (22% vs 14%).
Lastly, senior leaders lead on pride at 53%, versus 25% of owners/partners, who are the most likely to say nothing has made them avoid AI (53%).
As our last question, we asked US workers how they’d like to do their core work if handed a free choice. A combined 64% would either limit AI to small tasks or avoid it entirely, even at the cost of time.
A plurality (42%) would use AI only for small tasks and handle the important parts themselves. Another 22% would do the work themselves even if it took longer. Just 6% would use AI for everything they could, and the remaining 28% would use it to save time, even if it meant fixing mistakes.
Sex-wise, women lean toward minimal use, with 47% choosing AI for small tasks only, compared with 37% of men. Men are more likely to accept AI's mistakes to save time (33% vs 23%). Millennials most favor small-tasks-only, at 51%, compared with 33% of baby boomers.
Senior leaders (42%) and owners/partners (41%) prefer saving time even with fixes, versus 21% of individual contributors.