In an eight-month study of roughly 200 employees at a U.S. tech firm, researchers from Harvard found that AI tools tended to intensify work rather than reduce it. By accelerating task completion, AI enabled workers to take on a broader range of responsibilities—often extending their working hours in the process. Notably, although the company provided enterprise AI subscriptions, usage was voluntary. Employees adopted the tools on their own initiative.

Over time, however, once the novelty of these tools faded, many workers realized their workload had grown incrementally and almost imperceptibly. The researchers identified three primary drivers of this shift:

• AI made previously inaccessible tasks feel achievable—for example, enabling non-technical employees to engage in coding or engineering-related work.
• Lower friction in initiating and completing tasks blurred boundaries between work and personal time.
• AI facilitated seamless multitasking, functioning as a “background partner” capable of handling concurrent tasks—ultimately increasing overall task volume.

This study contributes to the broader debate on AI’s impact on employment. Some research emphasizes automation and job displacement; other perspectives highlight job creation. Harvard’s findings suggest a more nuanced outcome: AI may not simply replace roles or create new ones, but instead expand the scope of existing jobs—quietly adding more work.

Because the research focuses on a U.S. company, the results may also reflect elements of American workplace culture. Still, the study underscores a key tension: greater productivity often leads to heightened expectations. As AI enables individuals to accomplish more, they may feel compelled to do more—compounding existing concerns about displacement and job security. While every new technology demands adaptation, the challenge with AI may be learning how to set boundaries rather than continually increasing output.