New research suggests that while artificial intelligence promises to boost efficiency, it may unintentionally increase employee workload and intensify work demands rather than reduce them. According to a recent eWeek report, findings from an eight-month field study indicate that AI tools often create what the researchers call a “productivity trap,” where automation leads to more tasks rather than fewer.

The study, detailed in eWeek’s analysis of research published by the Harvard Business Review, tracked how generative AI changed the nature of work at a U.S. technology company. Rather than immediately alleviating routine tasks, employees found themselves managing additional responsibilities, handling more parallel work streams, and extending their engagement with work throughout the day.

Researchers describe this trend as “compound engineering”, a phenomenon where tools designed to streamline work end up adding layers of complexity: as workers experiment with AI and explore its capabilities, they accumulate new tasks that expand overall workload.

One of the study’s key takeaways is the paradox at the heart of AI adoption. As automation speeds up certain activities, expectations for output rise, which can lead employees to take on broader and more demanding roles. The research warns that this “workload creep” may ultimately lead to cognitive fatigue, burnout and diminished decision-making quality.

Such an increase in workload can lead to cognitive fatigue, burnout, and a weakened ability to make decisions,

the study states, underscoring the potential downside of unfettered AI use.

The implications extend beyond individual stress: companies could be inadvertently reshaping job expectations without formal guidance on how to manage AI tools responsibly. Experts cited in the report urge organizations to establish structured “AI practices”—clear norms for when and how employees should use AI—to prevent intensification from outweighing productivity gains.

As artificial intelligence becomes more ubiquitous in tech and business operations, this research adds to a growing dialogue about not only what AI can do, but how it should be integrated into workforces to sustainably support both efficiency and employee well-being.

Material by Yana Petrova

Image: Freepik