A workplace AI rollout can look successful if the people who already have the most access are the ones asked whether it helps. Gallup’s latest U.S. survey points to a more useful question for employers: who is using these tools, and whose work is changing as a result?
In findings published October 6, Gallup reported that 28% of U.S. employees used AI at work at least weekly, while 54% said they had not used it in their role. Weekly use was more common among college graduates than employees without a degree, and among managers than individual contributors. The survey does not show that AI created these gaps, but it does make them difficult to ignore when leaders assess a workplace rollout. (Gallup)
The comparison is not between a proven “AI workplace” and a proven alternative. Gallup did not test competing company policies. Instead, its results give leaders a way to compare two questions: should they focus first on workers already using AI, or on whether other employees have a fair chance to try it? The survey offers a reason to examine both, while leaving the best implementation strategy open.
What the gap looks like
Gallup found that 40% of college graduates used AI at work weekly, compared with 17% of employees without a college degree. The corresponding figures were 37% for managers and 25% for individual contributors. These are differences in reported use, not evidence that one group is more capable of using AI or that another has no useful tasks for it. (Gallup)
That distinction matters. A manager may have more discretion over tools or work processes than an individual contributor; a role may also have tasks that employees consider suitable for AI—or not. But the study, as reported, does not establish why these gaps exist. Treating them as proof of a particular cause would go beyond the findings.
The size of the nonuse group is also important for interpreting workplace enthusiasm. More than half of employees said they had not used AI in their role. That does not necessarily mean they were barred from it: the survey does not tell us whether nonuse reflected lack of access, lack of relevant tasks, a choice not to use it, or other reasons. It does mean that reports of benefits from users cannot be assumed to describe the experience of the whole workforce. (Gallup)
Two priorities, with different trade-offs
Prioritize current users. This approach starts with people already using AI and asks what they find useful. Among employees who had used AI at work, 63% said it helped them work faster, and 56% said it helped them find more creative solutions. Those responses can give employers a place to investigate potential uses and ask users for examples. But the figures are employees’ own assessments—not independent measurements of speed, accuracy or work quality. A leader who treats them as proof of productivity gains risks mistaking a positive perception for a demonstrated result. (Gallup)
Prioritize broader access and input. The other approach begins with the difference in reported use across education and job level, and asks whether people outside the more frequent-user groups have a meaningful opportunity to participate. That can make sense if leaders want to understand how the tool fits a wider range of roles. Its trade-off is that the survey does not identify which employees lack access, want access, or have tasks where AI would be useful. A blanket expansion would therefore be a decision to investigate or test—not a conclusion demanded by Gallup’s data. (Gallup)
These priorities need not be mutually exclusive. A company could learn from current users while also asking employees who have not used AI what accounts for that. That is a practical way to avoid making the most visible users the sole representatives of the workforce. It is a leadership recommendation based on the shape of the findings, not a specific policy evaluated by the survey.
What the survey can—and cannot—settle
The results come from the second year of the American Job Quality Study, a partnership involving Gallup, Jobs for the Future and the Families & Workers Fund. Gallup said the survey was conducted from January 26 through March 24, 2026. Its findings about AI’s effects drew on 7,845 employees who had used AI in their jobs at least a few times in the prior year. The survey was fielded months before its October publication, so it should not be read as a snapshot of workplace conditions on October 6. (Gallup)
Nor does the survey show that AI caused differences in job quality, or that AI use will create or eliminate jobs. Gallup notes that the groups reporting more AI use are also more likely to hold quality jobs, and says further research is needed to understand that relationship. That overlap leaves open multiple explanations; the findings do not settle which one is responsible. The study concerns U.S. employees, so its percentages should not be applied automatically to other countries or every occupation. (Gallup)
The findings also received attention in an October 8 HR Dive roundup, which highlighted Gallup’s results. That provides trade-press coverage, but the figures in this story come from Gallup’s study, not a separate survey or independent confirmation of the numbers. (HR Dive)
For a leader deciding what to do next, the most defensible choice is not to declare AI a success or failure based on these percentages. Use them to frame a workplace review: who is using the tools, who is not, and what do employees in each group say explains the difference? Then keep claims about benefits tied to what has actually been measured. Gallup’s data make uneven use visible; they do not prescribe a rollout or prove that wider use will improve work.
Sources
- AI Benefits at Work Unevenly Distributed (2026-10-06)
- What We're Reading (2026-10-08)
