British business owners are using artificial intelligence far less than their continental rivals, and a new survey of 4,510 founders across five European markets has put an uncomfortable number on the gap. Some 51% of UK founders said they had used AI to support their business in the past 12 months. In Germany the figure was 79%, in France 76% and in Spain 73%, according to the research, conducted by Pollfish and reported by IT Brief UK.
The gap between roughly half of British founders and four in five German ones is not small enough to be a survey artefact, and it is not obviously about capability. The UK has some of Europe’s strongest AI research institutions and government programmes aimed at exactly this adoption problem, including the DSIT research programme on AI adoption published earlier this year.
So what explains it? The survey points to cultural and educational friction rather than access. Founders who are sceptical of a tool until they see concrete value describe, fairly, what behavioural economists would call rational caution. But rational caution at the level of a market becomes a collective lag, and AI adoption tends to compound: firms that integrate the tools early learn which tasks they actually improve, which creates practice that slower movers then have to buy rather than build.
There is an admin angle that should worry policymakers more than the topline. The same European research found UK founders weighed down by tax and administrative burdens, an environment where the newest tools are often the ones that eat paperwork. If a fifth of a founder’s week goes on compliance, and AI handles chunks of compliance well, then the adoption gap is not just a technology story, it is a productivity one, in the country whose productivity puzzle has run for nearly two decades.
The counterweight is the flip side of caution. Britain’s regulators have moved early on AI governance questions, the ICO has published guidance on generative AI for businesses, and a survey finding of half adoption may partly reflect founders waiting for that guidance to settle rather than disinterest. German and French adoption figures may include plenty of experiments that never made it past a works council.
Either way, the number is now in the system, and it will be quoted in every briefing note written about the AI economy between now and the Budget. The interesting test will be the same survey next year: does the gap narrow because British founders moved, or because continental ones learned to be sceptical too?
This article draws on the Pollfish-commissioned founder survey reported by IT Brief UK, 16 September 2026, and GOV.UK’s AI Adoption research programme.