Despite their rapid adoption across white-collar industries, AI chatbots are saving workers just 3% of their time on average, according to new research from the National Bureau of Economic Research. The study also reveals that only a tiny fraction of these productivity gains, between 3% and 7% translate into higher pay for employees.
The research, conducted by economists Anders Humlum from the University of Chicago and Emilie Vestergaard from the University of Copenhagen, examined 25,000 workers across 7,000 workplaces in Denmark, focusing on professions thought to be most vulnerable to AI disruption such as accounting, customer support, legal work, and software development.
“AI chatbots have had no significant impact on earnings or recorded hours in any occupation,” the economists wrote in their working paper, challenging the narrative of AI as a workplace revolution.
Corporate Culture and the Untapped Potential of AI Productivity Gains
The findings come as a reality check amid spectacular claims about AI transforming workplaces. Companies like Duolingo have replaced contract workers with AI, while Shopify has famously declared it will only hire humans as a “second choice” to AI. Meanwhile, investors continue pouring money into AI-focused companies, driving valuations to extraordinary heights.
So why isn’t AI living up to the hype in real workplaces?
Part of the explanation lies in how employees use their AI-saved time. The study found that workers allocate more than 80% of time saved to other work tasks, including new responsibilities created by Artificial Intelligence itself like editing AI-generated content or adjusting systems to prevent AI cheating.
“I might save time drafting an email using a large language model, but the important question is, what do I use that time savings for?” Humlum explained. “Is the marginal task I’m shifting my work toward a productive task?”
Corporate culture also plays a significant role. Many employees use AI tools without explicit endorsement from management, limiting their ability to leverage productivity gains into better compensation or more meaningful work.
“In a workplace where it’s not explicitly encouraged, there’s limited space to go to your boss and say, ‘I’d like to take on more work because AI has made me more productive,'” Humlum noted.
The Reality of AI’s Productivity Impact
The research aligns with other recent indications that AI’s potential has been overstated. An IBM survey revealed that just 25% of AI projects deliver their promised return on investment, with corporate FOMO (fear of missing out) driving adoption more than clear business value.
Nobel laureate Daron Acemoglu estimates AI’s productivity boost at approximately 1.1% to 1.6% of GDP over the next decade significant but far from the economic doubling some tech enthusiasts predict.
“Getting productivity gains from any technology requires organizational adjustment, a range of complementary investments, and improvements in worker skills,” Acemoglu wrote for Fortune last year.
The Danish study confirms this perspective, showing better productivity when employers actively encourage the use of Artificial Intelligence and provide worker training.
Time may also be a factor. The Industrial Revolution transformed society over a century, not overnight. “It took a couple of decades to see that we can have an assembly line powered by electricity instead of having everything run centrally via a steam engine,” Humlum observed.
For now, the research suggests Artificial Intelligence is producing evolutionary rather than revolutionary change in most workplaces. Workers are finding modest time savings, but without corresponding improvements in compensation or dramatic shifts in work patterns.
As companies continue experimenting with AI integration, the real transformation may depend less on the technology itself and more on how organizations restructure work processes, train employees, and distribute the benefits of productivity gains.
