89% of Workers Now Use AI, But Most Have No Clear Rules for It

AI tools have gone mainstream in the American workplace. Almost nine out of ten workers (89%) say they’ve used AI tools for work and the majority of people use them on a regular basis. A new report by Founder Reports shows that 38% of workers said they use AI daily while another 23% said they use AI weekly. Just 11% said they‘ve never used AI tools to do their job.

And what’s even more significant is that many companies don’t have guidelines for how their employees use AI. The same poll of 2,078 employed Americans revealed that 44% said their employer doesn’t have an AI policy or they’re not sure if one exists. AI is rapidly gaining traction, but guidance is lacking.

AI Use Is Now the Default at Work

With 89% of workers using AI for their jobs and most using it at least weekly, the technology has moved past the early-adopter phase into everyday routine. 

But the overall numbers don’t tell the full story. There are some big differences in usage rates depending on roles and seniority.

Usage by department

Software engineers are by far the most likely to use AI daily (71%), ahead of marketing (63%), then IT and information security (55%). Conversely, departments with the lowest percentage of daily users are legal (25%) and operations (27%).

Usage by seniority

AI adoption also correlates with seniority. The percentage of C suite (62%) and vice presidents (63%) using AI on a daily basis is twice that for individual contributors (31%).

Most Companies Haven‘t Set the Rules

The most concerning finding from the survey is that a significant number of companies still do not have formal guidelines in place. The biggest single group of employees, 38%, said there was no clear AI policy at their workplace. 6% were not sure whether a policy exists. That means 44% of the workforce operates without knowing how or when they are supposed to use AI tools.

For companies that do have guidelines, the most common approach is to allow use with some restrictions (25%) . Just behind that is free use of AI with no restrictions (23%). Only 6% of workers say their employer states they have to use AI, and only 2% state that it’s forbidden.

Small companies are the most behind

The lack of clear policies is a bigger issue with small businesses. 59% of employees of firms with fewer than 10 workers said there is no policy or they do not know if one exists, compared to only 34% of those employed by firms with 1,000 or more workers. Mid-sized firms are in between, at about 43%. Smaller companies are making use of AI at a pace comparable to the large firms, but with much less structure around them.

Workers Trust AI-Assisted Work Less

Knowing that a coworker used AI changes how people view the output. Among those surveyed, 43% said they trust a colleague’s output less when AI was involved. That breaks down to 11% who trust it much less and 32% who trust it a little less. Only 20% said AI involvement made them trust the work more, and 37% said it made no difference.

Younger workers are the most skeptical

Workers under 40 were more skeptical of AI-supported work. 48% said that they trust it less than fully-human work, versus 34% of workers over 50.

Different job functions also hold different levels of trust: lawyers are the most untrusting, with 63% of those surveyed trusting AI work less, followed by design and creative (57%), then research (50%) and data analysis (48%).

AI Involvement Changes How Work Gets Reviewed

77% of respondents said they review a coworker‘s work “more carefully” if they know AI was used. 36% said they review AI work “much more carefully.” 

Even those who know AI tools best review output with more scrutiny. 80% of daily AI users will more thoroughly review a co-worker‘s AI-generated output. Being a power user doesn‘t seem to result in more trust toward the tools.

The skepticism is strongest in just a handful of functions, with design and creative the most likely to pay closer attention to AI work at 89%. Other functions where this tends to be the case include marketing (88%), research (86%), education (85%), and software engineering (84%). The full breakdown can be seen in the AI in the workplace stats Founder Reports published in conjunction with the report.

AI saves time for the author of the work, but extra review time is needed from anyone who receives the AI-generated work.

Fixing AI-Generated Work

The skepticism seems to be justified. Almost half of workers (45%) indicated that they had to redo or correct work done by one of their coworkers due to an overreliance on AI. Over half (59%) of daily AI users said they’ve had to fix AI work.

Perhaps the biggest surprise of the survey, however, was that managers are more likely than others to spend time clearing up AI slop. In fact, 57% of managers and above reported doing work related to AI cleanup versus only 38% of individual contributors.

Company policy impacts the experience too. At companies that mandate working with AI, 73% of employees have had to correct a coworker‘s AI-generated output, compared to the 38% at companies with no policy at all. Mandating the tools, yields more AI work, and more AI errors to spot.

A Workforce Moving Faster Than Its Rules

AI has become normal at work faster than companies have figured out how to manage it. Most people use the tools, but far fewer have a policy telling them how, and the downstream effect is more delays and rework.

That cost rarely shows up in a productivity reports, and data shows that the burden lands mostly on managers and senior leaders.

About the Survey

This report draws on a survey of 2,078 employed Americans, conducted through Prolific in April 2026. The sample was pre-screened to ensure respondents were U.S. citizens and currently working. The data was scrubbed to remove incomplete surveys. The report was created by Founder Reports, a business publication focused on entrepreneurship and the workplace.