OpenAI's economic research team maps how AI could reshape the EU workforce

OpenAI's economic research group has published an extension of its AI Jobs Transition Framework applied specifically to the European Union labor market, estimating how many EU jobs sit in occupations likely to grow, automate, or reorganize as AI capability spreads.

What's new

The report opens by drawing a distinction between how fast AI capability moves and how fast labor markets can absorb it: "AI capabilities can cross borders quickly. Jobs do not change in such a frictionless way." Applying its transition framework to EU employment data, OpenAI's researchers estimate:

• About 12% of EU employment is in occupations that may grow as AI lowers costs and expands the range of viable projects.
• About 14% is in occupations with relatively higher near-term automation potential.
• Another 27% is in occupations likely to be reorganized, where "AI may change workflows and skill needs" without necessarily displacing the role outright.
• The remaining 47% is in occupations facing less immediate change.

The report also finds meaningful country-level variation: Luxembourg, Sweden, and the Netherlands have larger shares of employment in occupations expected to grow with AI, while Germany, Greece, and Italy have larger shares in occupations with higher near-term automation potential. OpenAI notes that, compared with the U.S., the EU has a smaller share of employment in occupations with higher near-term automation potential.

Context

This extends a framework OpenAI has already applied to U.S. employment data in earlier releases of its economic research, now adapted to EU occupational and demographic statistics. It lands alongside a wave of similar labor-impact analyses from other labs and outlets this year, but is notable for coming directly from a frontier lab with visibility into actual usage patterns of its own models, rather than from an outside research group extrapolating from survey data alone. It also arrives as EU policymakers continue to work through AI Act implementation and broader workforce-transition planning, giving the report a direct policy audience.

Why it matters

The headline numbers are less important than the framing: OpenAI is explicitly telling policymakers not to rely on single aggregate statistics, and instead to plan at the occupation and country level, because the EU's automation exposure is unevenly distributed both geographically and by sector. That framing — reorganization affecting more workers (27%) than outright automation exposure (14%) — is a deliberate rhetorical move by OpenAI to counter narratives of mass job displacement in favor of a "workflows change, roles adapt" story, which matters for how the company positions itself in ongoing EU regulatory and labor-policy debates. For a company facing scrutiny in Europe over AI's labor-market effects, publishing granular, country-by-country estimates is also a way of shaping the terms of that debate before regulators or unions do it first.