According to a 2025 study by the University of Chicago’s Becker Friedman Institute, while generative AI tools like chatbots are saving workers modest amounts of time, these gains are often offset by the creation of new tasks, resulting in little overall impact on work hours or earnings. The research, based on large-scale surveys from 2023 and 2024, suggests that the anticipated sweeping labor market changes from AI adoption have yet to materialize.
AI Productivity Gains
Productivity boosts from generative AI remain modest, with workers reporting an average time savings of just 2.8% to 5.4% of their weekly hours-roughly 1 to 2 hours saved in a standard 40-hour workweek. These incremental gains have not translated into dramatic changes in workflow or job structure. Instead, the extra time is often funneled back into other work activities, rather than reducing overall hours or fundamentally altering how jobs are performed.
New Tasks from AI Adoption
A closer look at the 2023–2024 survey data reveals that about 8.4% of workers found themselves with entirely new responsibilities as a direct result of AI integration-regardless of whether they personally used AI tools. These fresh tasks often involve supervising or correcting AI outputs, adapting workflows to accommodate new tech, or handling exceptions where automation falls short. For some, the promise of AI-driven efficiency has come bundled with a to-do list that’s longer, not shorter, as organizations reconfigure roles and processes to fit around the limitations and quirks of their digital coworkers
Labor Market Impact
Despite the buzz, sweeping changes to wages and job security haven’t shown up in the data. Studies from late 2023 and 2024 consistently find no significant effect on overall earnings or hours worked, with even small shifts (greater than 1%) confidently ruled out by researchers. The modest productivity improvements and new tasks created by AI have, so far, failed to translate into either widespread job losses or pay increases, leaving the labor market looking surprisingly stable amid the AI hype
Study Conclusions
Researchers from the Becker Friedman Institute wrap up their findings with a clear message: the much-feared AI-driven labor upheaval hasn’t arrived-at least not yet. Confidence intervals in the data rule out even small effects (greater than 1%) on earnings or work hours, and the anticipated productivity windfall remains elusive. Any meaningful transformation in how organizations operate or how workers spend their days will likely require more time and deeper investment in integrating these new technologies into daily workflows
