AI Adoption Initially Hurts Productivity but Leads to Long-Term Gains, ECB Study Finds
A study presented at a European Central Bank (ECB) conference reveals that while artificial intelligence (AI) causes short-term disruptions for businesses, companies that survive the initial upheaval ultimately thrive. The research, which analyzed U.S. Census Bureau data and surveys from 2017 to 2021, found that early AI adopters in manufacturing initially experienced a drop in productivity as they replaced human workers with automation.
Contrary to the common belief that AI enhances productivity by augmenting jobs, the study suggests that the transition period can be painful. Kristina McElheran, a researcher at the University of Toronto and co-author of the study, explained, “In the short term, we see a lot of pain.” She attributed the decline to AI disrupting established business practices, such as inventory management.
However, firms that weathered the initial challenges saw significant long-term benefits, including higher sales growth, improved productivity, and increased employment. “Surviving this seems like part of the problem,” McElheran noted, emphasizing that only adaptable companies reaped the rewards. Older, larger corporations often struggled to adjust, limiting their ability to capitalize on AI’s potential.
The study examined 30,000 firms, with AI adoption rates rising from 7.5% to 9.1% during the research period. ECB President Christine Lagarde, who introduced the conference, acknowledged that 23% to 29% of European workers are highly exposed to AI-driven changes. However, she dismissed fears of a “job apocalypse,” predicting that while some roles may disappear, new opportunities will emerge.
The findings highlight a critical lesson for businesses: short-term AI disruptions may lead to long-term success—but only for those that adapt effectively.
