How Europe’s Transformative AI Strategy Aims to Avoid Dependence

Europe should treat advanced AI as strategic economic and security infrastructure—not as a technology it can safely obtain on demand from foreign providers. That is the central argument of A Transformative AI Strategy for Europe, which warns that, without urgent action, Europe risks marginalisation and dependency as AI capabilities advance. 16

The headline goal: build far more AI compute in Europe

The report’s most visible recommendation is a large-scale expansion of AI data-centre capacity. It calls for Europe to host 15% of global AI compute on EU soil by 2030, compared with roughly 5% today. The strategy’s proposals have been reported as requiring a €100 billion public commitment for data centres and compute, alongside an ambition to mobilise €1.5 trillion in private investment. 356

The purpose is not infrastructure for its own sake. The authors see domestic compute capacity as a foundation for Europe’s prosperity, sovereignty and security: without it, European companies, researchers and governments could remain reliant on infrastructure controlled elsewhere. 1618

Use public funding to unlock a wider build-out

The strategy envisions public intervention that reduces the risk of strategically important private data-centre projects. Its published recommendations include €75 billion of national public funding through instruments such as loan guarantees, advance-purchase commitments and possible equity investment; reporting on the broader strategy describes a €100 billion public push and a €1.5 trillion private-capital objective. 36

That distinction matters. The proposal is not framed as the state building every facility itself. It is an industrial-finance strategy intended to make enough European compute capacity investable and available at scale.

Secure access to frontier models while building European capability

The report does not assume Europe can immediately replace every leading foreign AI provider. Instead, it argues Europe should secure reliable frontier-AI access—including models and compute—while strengthening its own technical and institutional capacity.

The authors state that Europe is currently dependent on foreign providers for frontier-AI access. They recommend using Europe’s technological and economic assets to negotiate guaranteed access, while protecting those assets from foreign acquisition or loss of value. 1718

In practice, this makes the strategy a two-track approach: expand European infrastructure and expertise, while ensuring Europe is not cut off from the frontier during the transition.

Give EU institutions the technical capacity to act

The report also says Member States and EU institutions need frontier-AI expertise and technology inside government. The aim is for public institutions to keep pace with developments rather than depend entirely on outside companies to understand, procure, assess or respond to advanced AI systems. 12

This is a governance recommendation as much as a technology one. A government cannot effectively oversee or use strategically important AI if it lacks the capability to evaluate what the systems can do and what dependencies they create.

Pool supply-chain leverage across Member States

Another central recommendation is a Member State Alliance for Supply Chain Security. The report argues that European countries possess important chokepoints in the global AI supply chain but do not fully use their collective strategic and economic leverage. 1217

A coalition of willing Member States, coordinated with the European Commission, could pool that leverage in negotiations and security planning. The report presents this as a way to improve collective bargaining power and make frontier-AI access more reliable for Europe as a whole. 18

Plan for AI-related disruption, not only economic competition

The strategy frames AI as a resilience and preparedness issue as well as a competitiveness issue. Its recommendations include making European institutions ready to act and ensuring resilience to the risks associated with transformative AI. 1216

The wider implication is that crisis planning should account for AI-amplified threats alongside the opportunity to use advanced systems in public administration and economic activity. The report’s case is that access, infrastructure, institutional competence and security planning are interdependent—not separate policy files.

The bottom line

The report’s proposed answer to dependence is a coordinated European AI industrial-and-security strategy:

• expand AI compute on European soil toward a 15% global share by 2030;
• use public commitments to mobilise far larger private investment;
• secure dependable access to frontier models and compute while developing domestic capability;
• put frontier-AI expertise inside EU and national institutions; and
• pool Member State leverage over strategic AI supply chains. 361218

Its core claim is straightforward: Europe cannot preserve meaningful choice over its economic and security future through regulation alone. It must also possess infrastructure, technical capacity and bargaining power in the AI ecosystem. 1618