Are AI Chatbots Truly Transforming the Workplace?

Researchers Anders Humlum (University of Chicago) and Emilie Vestergaard (University of Copenhagen) conducted an in-depth study, analyzing responses from two large surveys (2023 and 2024) alongside detailed job and pay records in Denmark. Their focus: 25,000 employees across 7,000 workplaces in 11 roles deemed susceptible to AI disruption.

AI Chatbots Are Everywhere, But Where’s the Impact?

The findings confirm that AI chatbots have become ubiquitous in Danish workplaces. Employers, once wary of data privacy concerns, are now actively encouraging their use. Nearly 40% of companies have implemented in-house chatbots, and a third of employees have received formal training. This shift has significantly increased adoption rates, with usage jumping from 47% to 83% when employers endorsed the tools.

Interestingly, this encouragement also narrowed the gender gap in chatbot usage, particularly when paired with training. Yet, despite widespread adoption, the anticipated economic benefits remain elusive.

Using statistical methods, the researchers found no significant changes in pay, work hours, or job retention across the 11 roles studied. Even among early adopters or workplaces with high chatbot usage, there were no notable differences in hiring trends, wages, or staff retention.

Productivity Gains: Incremental, Not Transformative

Why hasn’t the hype translated into measurable outcomes? The study points to modest productivity gains and their limited impact on wages. While users reported improved work quality, creativity, and time savings, the average time saved amounted to just 2.8% of total work hours—far below the 15%-50% gains seen in controlled experiments.

The discrepancy lies in the broader scope of this study, which included roles like teaching, where chatbot benefits are less pronounced. Controlled experiments often focus on tasks where AI excels, such as coding or customer service.

The researchers emphasize the importance of “complementary investments.” Companies that provided training and fostered a supportive environment saw greater benefits, suggesting that simply deploying the tools isn’t enough to unlock their full potential.

Even so, these modest gains haven’t translated into higher earnings. Only 3%-7% of the time saved appeared as increased wages, possibly due to workplace inertia or the informal nature of chatbot adoption.

New Tasks, Not Just Faster Work

AI chatbots are also creating new tasks. About 17% of users reported additional workloads, often involving integrating AI into workflows, drafting content, or addressing ethical concerns. Interestingly, even non-users felt the ripple effects, with 5% reporting new tasks like adapting assignments or detecting AI-generated work.

This trend was more pronounced in workplaces that encouraged chatbot use, highlighting the ongoing adaptation phase as companies figure out how to best leverage the technology.

The Verdict: A Slow-Burning Revolution

The researchers caution against dismissing generative AI’s potential. While its current impact on productivity and labor market outcomes is minimal, they see room for growth as companies refine their integration strategies and as new tasks evolve.

For now, however, the reality falls short of the hype. Two years after ChatGPT’s launch, the transformative impact on jobs and pay remains surprisingly modest. The revolution may still be on the horizon, but it’s unfolding at a slower pace than anticipated.