A Hubstaff-commissioned poll of 1,068 full-time U.S. workers in computer-based jobs found 46% of AI users said they were expected to handle more after starting to use AI, while 26% applied saved time to extra work.
Hubstaff hired Pollfish to survey employees ages 18 to 64 in October. Overall, 79% said they used AI for work. Among AI users who reported time savings, 76% estimated they gained at least an hour in a typical week, including 32% who put the savings at four hours or more.
Asked what they did with the time, 39% said they finished work earlier, 26% put it toward additional tasks and 15% said little changed. On workload expectations, 5% reported a decrease after they began using AI.
The poll also captured the time workers spend checking colleagues’ AI output. Fifty-three percent of AI users said they received work needing substantial correction at least once a month, while 45% said they spent an hour or more each week reviewing or fixing such material. The results describe separate reported effects; they do not measure the net impact across an organization.
Hubstaff CEO Jared Brown said, “AI is giving time back to individual employees, but most organizations aren't capturing it as capacity.” Brown pointed to employer capacity planning for 2027 as a possible use for the findings.
A separate analysis by the U.S. Census Bureau used responses from March 2026. Among workers who had used AI on the job during the prior week, 31% estimated savings of one to two hours, 15% put savings at three to four hours and another 15% said they exceeded four hours. The Census survey covered 11 workplace tasks and asked how much extra time respondents would have needed without AI, so it uses a different measure from Hubstaff’s typical-week estimate.
Hubstaff reported whole-number percentages and said it did not weight the results. It gives the full sample a 95% confidence range of plus or minus 3 percentage points, with wider margins for smaller groups. The poll records workers’ accounts; it did not track company output or determine whether AI use drove changes in expected workloads.
