Artificial intelligence is taking on more than repetitive tasks. New research from Atlassian finds that employees are handing over work they enjoy at nearly the same rate as work they dislike, including tasks that help them build expertise.
Among AI users, 82% said the technology handles at least some of the work they enjoy and 82% said it handles at least some of the work they dislike. Ten percent said AI has taken over all of their enjoyable work, compared with just 3% who said the same about tasks they dislike.
Heavy AI Use Means Less Employee Ownership Of Enjoyable Work
The more deeply AI is embedded in team workflows, the more enjoyable work employees hand over to it. Among occasional users, 24% said AI handles a moderate or greater share of the work they enjoy. That figure rises to 71% among employees whose teams have made AI a standard part of their workflows.
The trend is also pronounced among senior employees. Sixty percent of directors and above said AI handles a moderate or greater share of their enjoyable work, compared with 42% of managers and 36% of individual contributors.
This raises concerns about how less experienced employees will develop skills if AI takes over tasks that traditionally provide opportunities to learn.
AI Makes Specialized Work Easier, But Expertise Still Matters
AI is helping employees without deep expertise take on more specialized tasks. Around 60% of users said it makes analyzing information and reviewing or improving work easier for non-experts, while 59% said it helps them produce first drafts.
However, only 28% believe AI allows non-experts to make decisions that previously required a specialist. More than half of AI users, 54%, expect the technology to create more specialists because expert judgment will remain important for guiding work and catching mistakes.
Employees Want Humans In Charge Of Important Decisions
Workers are generally comfortable with AI handling routine tasks, particularly reviewing large volumes of data and carrying out repetitive actions. But they want people to lead when decisions require judgment.
Sixty-six percent want humans to determine priorities, 67% want human leadership on deliverables requiring deep expertise, and 78% want people in charge of high-risk decisions. Just 3% would let AI make high-risk decisions independently.
