Can Europe have its AI moment?

Yesterday, Mistral AI came to Brussels with Arthur Mensch to present a new white paper, “European AI: a playbook to own it” - an evening focused on how Europe can actually lead in AI.

The message wasn’t that Europe is behind. It’s that Europe isn’t yet acting like it fully understands what it already has: world-class universities, a 450-million-person single market, and a strong commitment to human-centered technology. As colleagues at the Centre for European Policy Studies have argued, the core ingredients are there - the constraint is not capability, but unity and decisive action.

Mistral’s playbook is concrete, actionable and drawing from experience in the field: fast-track AI talent visas, build joint academic–industry PhD programs, create a unified digital procurement gateway, and invest in sovereign compute infrastructure for frontier AI. The ambition goes further toward a real European stack: cloud, data commons, and simpler regulation that avoids fragmentation across 27 rulebooks. Bold, unified, fast - three words not often associated with EU policymaking, but yesterday they felt within reach.

One issue, however, remains under-addressed - the elephant in the room: capital. In 2025 European startups captured only 5.13% of global venture funding, and institutional investors allocated a tiny amount of their assets to AI and deep tech. The playbook is strong and needs to move quickly - but without unlocking private capital at scale, Europe risks designing a strategy it cannot afford to execute.