On 3 June, the European Commission unveiled its European Technological Sovereignty Package — a Chips Act 2.0, a Cloud and AI Development Act, an open-source strategy, and a roadmap for digitalisation in energy. The stated ambition is striking: Europe should take control of its own data, supply chains, and future.

The impulse is understandable. Europe remains heavily dependent on non-EU suppliers for core digital technologies, and demand for compute is rising sharply. In a world where the US and China are each pursuing full-stack AI dominance, drawing other economies into their orbits, Europe risks being included in the AI order on terms set by others.

The Commission deserves credit for combining regulatory ambition with industrial policy at last. But the package conflates two distinct things: owning the stack and controlling it. Until Europe grasps that distinction, its third way will remain more declaration than architecture.

The AI race is largely a race for compute. Whoever controls that infrastructure shapes the technology and sets the terms on which others access it. Europe is not a meaningful player: it produces a handful of models of limited global reach, attracts only a fraction of global AI investment, and routes most sensitive workloads through American hyperscalers.

What makes this doubly frustrating is that Europe holds upstream hardware assets – a global monopoly in extreme ultraviolet lithography through ASML, and frontier semiconductor research through Belgium’s IMEC – yet exports these advantages to the US, South Korea, and Taiwan.

The Chips Act 2.0 addresses this in part, but building competitive fabrication capacity takes a decade. The Cloud and AI Development Act introduces a useful sovereignty assessment framework, but assessment is not infrastructure.

The most instructive comparator is China. Unlike the US, which has been at the frontier throughout, China started from dependency on foreign hardware and responded with a two-track strategy: using diplomatic and market leverage to extract concessions from Washington, while building a domestic ecosystem around Huawei’s Ascend chips and state-directed procurement.

Chinese firms unable to source Nvidia hardware now rely on Ascend for a large share of their compute. That is what closing a hardware gap from dependency looks like — but it will be neither fast nor cheap.

The harder comparison is India. India’s digital public infrastructure, such as the account aggregator framework, was a genuine innovation: population-scale identity and payments rails that Europe has studied admiringly. But AI sovereignty has proved far more demanding. India imports virtually all its advanced semiconductors, depends on foreign hyperscalers for cloud inference, and lacks the energy infrastructure for large-scale GPU clusters.

The India Stack showed that building public-purpose digital rails is tractable; controlling the full AI stack is not, even for a country with strong software talent, 1.4 billion users, and substantial political will. India has struggled partly because it pursued sovereignty across the full stack rather than focusing control on the layers that most directly affect strategic exposure. The result has been diffuse effort and limited leverage anywhere.

The right reframing for Europe is therefore not “can we own the stack?” (the answer is almost certainly no, not in full, not soon), but “can we control how it behaves within European society and close our most critical hardware dependencies?” Those are more tractable questions.

Control means embedding compliance, audit trails, and risk thresholds into procurement contracts as conditions of market access, not afterthoughts. It means building genuine evaluation capacity so that European public authorities can test AI systems for reliability, security, and bias rather than relying on vendor assurances. It means using Europe’s single market as leverage — requiring data residency, algorithmic auditability, open interfaces, and sovereign fallback options from any firm operating at scale in Europe.

On hardware specifically, it means treating ASML and IMEC not merely as export champions but as anchors for a European semiconductor industrial policy — using public procurement, AI Gigafactories, and the EIB to create the captive demand that China’s state-directed procurement created for Huawei.

Without a credible domestic customer base, the infrastructure investment in the Commission’s package will struggle to become self-sustaining.

The Commission’s Technological Sovereignty Package is a start. What it needs to become is a control architecture grounded in honest industrial strategy – one that learns from China’s hardware trajectory, avoids India’s diffusion trap, and uses Europe’s regulatory market power not as a substitute for infrastructure, but as the lever that makes infrastructure investment worthwhile.