No European labour market is fully safe from AI-driven job displacement.
We stress-tested 36 markets across five lenses and eight scenarios. Most countries can only handle some of the disruption, with 15 probably already being past their limit. Under the rules we applied, only nine have the statistical strength to hold up under pressure.
“Will AI kill all jobs as we know them? No. It will force us to renegotiate what a job is and what it’s worth.”
AI won’t eliminate work. It will reshape many roles, create new ones, and force a societal, political, and economic negotiation over what we count as “work” and how we value it.
“Over 75% of Europe’s labour markets face a rocky future.”
Most countries can absorb only part of an AI disruption. Fifteen look as if they are already beyond what they can handle, and what counts as “safe” depends on the rule you use (9 are safe under a semi-flexible rule, but zero under a strict one). Even Norway and Sweden, the two most stable countries, fail the strict rule. The usual “safety valves” people expect to ease AI job disruption (older workers retiring and others retraining) are not moving fast enough to keep pace with how quickly jobs could change.
Go deeper →“Eight storylines for the next decade, with the odds shaped by each country’s economic regime.”
We tested seven routine futures plus the “special” one (Polycrisis Drag), where several shocks happen simultaneously. The odds for each future shift depending on whether the economy is growing, stuck, or shrinking. For example, in a shrinking or non-growing economy (“post-growth”), the usual “tech boom brings jobs back” outcome becomes much less likely (about 20% → 13%), while a “jobs shift into climate adaptation work” outcome becomes more probable.
Go deeper →“Europe doesn’t have one answer. It has 36.”
On a population-weighted average, the EU-27 looks as if it is coping. But the country-by-country view tells a different story. Even if the worst knock-on effects of AI-driven job disruption build up mainly in countries outside the EU, no EU-27 or EFTA country is safe. For about 40% of workers, the “best-case” path depends on climate-adaptation jobs rather than a tech-led rebound.
Go deeper →“How we tested and what we deliberately left open.”
We chose these five lenses because earlier parts of the project had already gathered the evidence, not because they are the only ones. Overall, this analysis has three known gaps we could not measure well. We also considered, but left out, other possible future storylines because we lacked solid real-world data to justify including them.
Go deeper →“Prefer it as a document? Four PDFs cover the full analysis.”
The same data, five lenses, eight scenarios, four fragility classes, packaged four ways. Visual Read for a meeting, Executive for a briefing, Long-Read for the full argument, Specialist for the data.
Get the PDFs →