Artificial intelligence is not yet shortening the workday — at least not for those who actively use it. Workers who use AI in their work spend an average of 8.4 to 9.1 hours per day on tasks. For those who do not use such tools, the figure is 7.6 hours.
Artificial intelligence is not yet shortening the workday — at least not for those who actively use it. Workers who use AI in their work spend an average of 8.4 to 9.1 hours per day on tasks. For those who do not use such tools, the figure is 7.6 hours.
This is evidenced by an analysis by ActivTrak, a company that develops software for assessing labor productivity.
The study covered the digital activities of more than 9,000 employees from 467 companies between April and September 2026. According to the analysis, the working day of AI users can be up to 90 minutes longer than those who do not use these tools.
The biggest difference is among active users, defined by ActivTrak as employees who use AI for more than 45 minutes in at least five sessions per day. They also spend an average of 18 minutes more on their work after work than non-AI users.
Additionally, active users switch between applications more often and have shorter periods of continuous focused work than those who use AI less frequently.
“AI can help workers complete tasks faster, but our research shows that it also contributes to longer workdays, more frequent switching between applications, and shorter periods of concentration,” said Gabriela Mauch, Chief Customer Officer and Head of ActivTrak Productivity Lab.
The researchers offer several explanations for the results. It could be that people who were already working longer hours were more likely to use AI. Another possibility is that AI actually encourages workers to take on more tasks. It could also be that the extra time is spent reviewing and correcting the results generated by the AI.
“We live in an economy where productivity is extremely important: if we get more of it, we want even more,” Mauch explained in a comment to Bloomberg.
Similar findings were made by researchers at the Haas School of Business at the University of California, Berkeley. In February 2026, they published the results of an eight-month ethnographic study at a technology company with a staff of about 200 people.
Researchers found that using AI didn’t necessarily free up workers' time. Instead, they started work faster, took on a wider range of tasks, and worked longer hours—often without direct requests from management.
One of the study’s authors, Xinzi Maggie Ye, warned that this model could be unsustainable in the long term. As the number of tasks and parallel AI processes increases, the workday becomes denser and more cognitively complex. Gradually, the additional workload that workers initially voluntarily take on can become the new norm.
