Aruna Ranganathan, an assistant professor of management at the Haas School of Business at UC Berkeley, and Maggie Sinzi, a graduate student from the same school, came to this conclusion.
Their eight-month study was conducted at a U.S. technology company with about 200 employees. During this time, the authors observed teams of engineers, designers, product managers and operations staff, monitored internal communication channels and conducted more than 40 in-depth interviews, writes The insider.
What the authors of the study revealed
The researchers identified three forms of labor intensification.
Thefirst is task augmentation: AI fills knowledge gaps, and workers began to take on responsibilities that were previously performed by colleagues. Product managers and designers started writing code, researchers took on engineering tasks.
The second form is the blurring of the boundaries between work and personal time: a low threshold for entering a task encouraged employees to send AI requests during lunch, breaks, or late at night, without perceiving it as a full-fledged job.
Third, the rise of multitasking: running multiple agents in parallel and constantly switching between threads created a continuous cognitive load.
The result was a self-reinforcing cycle. AI accelerated tasks, which raised expectations of speed; higher expectations increased dependence on AI; dependence widened the range of tasks, and widening the range increased the density of work.
One engineer described this paradox this way. “It would seem that since AI increases productivity, it means you save time and work less. But in reality, you don’t work less. You work the same amount or even more.”
Burnout is a “side effect”
Ranganathan and Sinzi warn that short-term increases in productivity can mask increasing overload, which over time translates into cognitive exhaustion, burnout, and decreased decision quality.
As a counterbalance, they suggest that companies form what they call “AI practices” – a set of conscious norms and procedures governing the use of the tools.
Specific measures include instituting mandatory pauses before making important decisions, organizing notifications into scheduled “packets” to reduce fragmentation of employee attention, and setting aside time for live interaction with colleagues to restore insight and support creative thinking.
Last December, it found that the percentage of U.S. employees who use artificial intelligence technology in their professional lives at least a few times a year rose from 40% to 45% between the second and third quarters of 2025.
