Study finds AI tools expand employee workloads
A new Harvard research study challenges the narrative that AI tools primarily boost productivity by automating simple tasks. The study found that AI is often expanding employee workloads, as users take on more tasks and work longer hours, sometimes leading to burnout.
- The Harvard Business Review study involved an eight-month observation of approximately 200 employees at a U.S.-based tech company. Researchers Aruna Ranganathan and Xingqi Maggie Ye found that employees voluntarily using AI tools worked faster, took on a wider range of tasks, and worked longer hours. - This phenomenon is termed "workload creep," where the use of AI blurs the boundaries between work and personal time, leading to cognitive fatigue and potential burnout. The study noted that employees began using AI during breaks, at night, and in the early mornings because the tools made it easier to start and continue tasks. - The research identified a trend of "task expansion," where employees in roles like product management and design began writing code, and researchers took on engineering tasks. This, in turn, created more work for software engineers who then had to review code from colleagues with less experience and act as mentors. - This study contributes to the broader discussion of the "AI productivity paradox," where individual productivity gains from AI don't always translate to measurable improvements in organizational performance or higher pay for employees. While over 75% of developers report using AI coding assistants, many companies are not seeing a corresponding increase in delivery velocity. - One in five employees have encountered misinformation or errors from AI tools, requiring manual correction that can slow down projects. The initial surge in productivity from AI can be followed by a decline in work quality and increased employee turnover if not managed. - A Federal Reserve Bank of St. Louis report from February 2025 indicated that workers using generative AI saved an average of 5.4% of their work hours. However, only 3-7% of these productivity gains translate into higher earnings for workers. - The increased workload isn't always a result of direct managerial requests. The study found that employees were intrinsically motivated to take on more tasks because AI made previously daunting or outsourced work feel more achievable. - To mitigate the negative effects, researchers suggest that companies need to establish clear guidelines for AI use to prevent employee burnout. This includes regulating the timing and order of work to reduce cognitive overload and preserving time for human interaction.
