EU Frontier AI Sovereignty: Beware of Geeks Bearing Gifts

Europe’s most consequential AI dependencies may not arrive through coercion, but as gifts: infrastructure investments, model access agreements, or technology partnerships that appear beneficial in the short run but quietly erode Europe’s capacity for self-determination. Our new report, “Beware of Geeks Bearing Gifts: Building True EU Frontier AI Sovereignty”, provides a framework to assess the trade-offs across sovereignty pillars and frontier AI stack components.

Why EU Frontier AI Sovereignty Matters

Frontier AI is not just another technology sector. Unlike prior general-purpose technologies, it is designed to act with increasing autonomy, embedding the values and priorities of its developers into everything it touches. With a projected impact on global economic output of up to 15% over the next decade, it is already reshaping labour markets, democratic processes, and the conduct of warfare.

However, Europe enters this transformation from a position of structural dependence: frontier AI models are almost exclusively produced in the US and China, and Europe’s supercomputing capacity relative to global leaders is in consistent decline.

Europe’s Current AI Strategy: Progress, but Fragmented

Following the call to action in the 2024 Draghi report, the European Commission launched an array of AI initiatives spanning compute infrastructure, data governance, regulatory reform, research funding, and skills.

Yet our analysis of 92 initiatives across four major communications shows that these initiatives not only remain disconnected but also disproportionately address sovereignty as an issue of economic competitiveness.

A New Framework for EU AI Sovereignty

With this report, we reframe the sovereignty debate. First, we expand sovereignty into five distinct pillars that, together, define what it means for Europe to remain a self-determined actor in the age of frontier AI. Second, we map the frontier AI stack at a finer level of granularity than most prior analyses. Third, we unite the two into a single EU AI sovereignty framework that lets policymakers design a more coherent EU frontier AI sovereignty strategy.

The Five Pillars of EU Frontier AI Sovereignty

Figure 1. Five pillars of sovereignty.

With this categorisation, we make explicit what current policy debates leave implicit: advancing one pillar can come at the direct expense of another. We do not seek to prevent these trade-offs—for they are inherent to the global nature of frontier AI. Instead, we seek to make them visible, so that decision-makers can navigate them deliberatively rather than stumble into them.

Mapping the Frontier AI Stack

We also provide a comprehensive map of the frontier AI stack. Frontier AI is not a single technology but the result of many interacting layers. We decompose this stack into five layers, 26 components, and 29 sub-components—a finer level of granularity than most prior analyses—to help policymakers identify which layers require intervention, which can serve as strategic leverage points, and where Europe holds structural strengths or critical vulnerabilities.

A Unified EU AI Sovereignty Framework

Together, these two contributions result in a unified framework connecting the five sovereignty pillars to the full frontier AI stack. To showcase its potential, we apply the framework to five EU initiatives (including AI regulatory sandboxes, InvestAI, and the European Health Data Space) and conduct a deeper analysis of the EU’s flagship AI Gigafactory Initiative. The latter reveals how, under certain circumstances related to ownership, access, or governance, an attempt to close Europe’s compute gap could deepen, rather than reduce, foreign dependencies.

From Crisis Mode to Strategic Leadership

Our mapping of 92 initiatives shows that the means and ambitions are present across the frontier AI stack. What is needed now is a clear and pragmatic direction to address the challenges that frontier AI is already posing. The EU has emerged structurally stronger from prior crises, and frontier AI must be no different. This, however, will require treating AI sovereignty as the multidimensional strategic challenge it is. 

The question Europe faces is no longer whether frontier AI sovereignty matters, but whether the EU and its member states will act with sufficient strategic ambition to shape their own future.

Read the full report for the complete analysis of EU frontier AI sovereignty, Europe’s AI dependencies, and strategic policy pathways forward.

For more insights, research, and expert analysis, follow The Future Society on Linkedin.