New research out today: Where is #GenAI actually impacting labor markets?
As part of our World Bank work supporting countries with their AI policies and strategies, an important element is to understand the potential impacts of generative AI in labor markets. Yet, most studies on GenAI and jobs rely on expert estimates and projections.
Working with Anna Powell-Smith (and insights from Luke Jordan), we took a different route: using new real-world data from Anthropic’s Economic Index to estimate the localized impact of AI, mapping usage patterns to employment across 33 European economies and U.S. counties.
The results suggest some long-held assumptions may not hold. These include:
1) It’s not repetitive, routine jobs being hit first: it’s high-skilled, professional roles
2) Wealth doesn’t predict adoption (e.g. Ireland lags Denmark)
3) Infrastructure isn’t destiny (e.g. Spain has near-universal broadband but relative low AI usage)
4) In Europe, Netherlands, Sweden, Denmark, and Finland lead in likely GenAI labor market impact
5) In the U.S., GenAI use isn’t confined to Silicon Valley or large metropolitan areas: for instance, parts of Colorado and Texas show surprising intensity
The “last mile” of GenAI adoption may be professional, not physical, shaped more by occupational structure than digital infrastructure alone.
This has real policy implications. High-usage economies should prepare for professional transformation through reskilling and responsible AI integration.
Lower-usage regions risk being left behind if they don’t act strategically. The future of work with GenAI will not be evenly distributed. But it is already here. And it is beginning at the top.
Link to the original analysis in the comments 👇