Highly engaged employees spend one and a half weeks per year on average fixing mistakes made by artificial intelligence (AI), a new survey has found.
Many enterprises adopt the technology in hopes of increased productivity. However, a recent Workday survey suggests that productivity gains alone don’t translate into better outcomes for most organizations.
The findings reveal that AI is helping employees complete tasks faster – most report saving between one and seven hours per week.
However, employees spend significant amounts of time correcting or rewriting low-quality AI-generated content. As a result, about 37% of the time saved by AI is wasted on redoing the work.
Nearly eight in ten (77%) employees report that AI work is audited with the same or greater rigor than human work. The trend is particularly prominent among employees who use AI daily.
While most (66%) business leaders consider AI skills training a top priority, fewer than half (37%) of employees who redo AI work the most reported increased access to training.
The survey was conducted in November 2025 and included 3,200 respondents who currently use or are personally exposed to AI at work.
All of them were required to work in a company with at least $100 million in annual revenue and have 150 or more employees.
AI productivity may be overhyped
Despite business leaders’ hype about becoming an AI-first company, employees who use the technology for daily tasks offer a more sobering view.
As many as 56% of Australian STEM employees say AI only increases their workload, according to a recent survey.
For instance, using AI copilots helps to write more code, but it also results in more code being reverted, removed, or updated.
At the same time, data suggests that AI is already capable of replacing nearly 12% of the US workforce, with junior or entry-level roles being the most vulnerable to AI adoption.
Unlock exclusive Cybernews content on YouTube.
Your email address will not be published. Required fields are markedmarked