The central argument of this report is simple: European ‘digital sovereignty’ is a grand delusion. It is delusional politically, because sovereignty belongs to sovereign nation-states and the EU is not one. It is delusional empirically, because no actor in the world – including the United States and China – possesses sovereignty across the full digital stack. And it is delusional strategically, because the EU’s preferred instruments – from regulation through champion designation and funding announcements to choosing champions and setting targets – have weakened the very conditions from which genuine technological power would have to grow. What Europe needs is not digital sovereignty. It needs agency, leverage and indispensability.
The modern digital economy is not a territory to be enclosed. It is a layered, global, interdependent network of platforms, supply chains, standards, research infrastructures, manufacturing capabilities and irreplace able technical nodes. The United States dominates cloud, operating systems, browsers, frontier AI and chip design, but remains dependent on Taiwan for leading-edge fabrication and on ASML in the Netherlands for EUV lithography. China has built the most extensive apparatus of domestic digital control in the world, but cannot manufacture the most advanced chips at scale because it lacks access to ASML’s EUV machines. The EU controls neither the platform layer nor the AI layer nor the cloud layer nor the end-user operating systems. Its distinctive power lies in regulation. But regulatory power is not digital sovereignty; it is the power to regulate technologies built by others. Ironically, the EU’s regulatory sovereignty depends on the continued presence of American and Chinese firms in Europe, while its rules make it harder for European firms to become their challengers.
This report illustrates the problem of the EU’s regulatory mindset through two contrasting case studies. Minitel was the closest historical approximation to digital sovereignty in practice. France built, owned and controlled a national online infrastructure in the 1980s, years before the web. While it was technically advanced, commercially successful and socially embedded, it was closed, nationally controlled and monetised through France Telecom’s monopoly. When the open internet arrived, the sovereign system became an obstruction. France, having pioneered mass-market online services, entered the internet age late and produced no global digital platform. Minitel shows that territorial control of a digital system can be achieved only at the cost of openness, scale and adaptability – the very qualities that made the internet dominant.
ASML shows the opposite path. It was not created by an EU digital- sovereignty programme. Its power rests not on owning a bounded territory, but on occupying an indispensable node in a global system. Every advanced semiconductor manufacturer depends on ASML’s EUV lithography machines. ASML is powerful because the world must pass through it. That is not sovereignty. It is chokepoint power that gives Europe global leverage.
If one doubts the critical importance of chokepoints in geopolitics, one only has to examine the war in Iran today and the role of the Strait of Hormuz. This may be in the analogue world, but the Strait’s strategic importance, and the power those who control it can command, reveals a sharp lesson for the digital world, too.
This is the report’s core strategic lesson. In a networked world, power does not come from trying to replicate the entire stack behind a European wall. It comes from identifying and cultivating nodes without which the system cannot function. Chokepoints confer agency: the capacity to shape outcomes because others depend on your capability. Ironically, it is the USA, not the EU, that has deployed ASML’s chokepoint power as part of its geopolitical interests. Washington has used ASML to progressively restrict China’s access to advanced semiconductor manufacturing capability. The EU’s digital-sovereignty discourse performs ambition, but ASML demon strates where the actual power lies in the twenty-first century.
Europe does have assets from which such indispensability can be cultivated: IMEC in semiconductor research, Airbus in large civil aviation, Vestas and Siemens Gamesa in wind, Fraunhofer in applied research, Novo Nordisk in GLP-1 therapeutics, SAP in enterprise software, and European strengths in photonics, quantum, industrial software, precision biotech and AI evaluation. None is a perfect analogue to ASML. But they show Europe’s potential.
Where European institutions create conditions and step back, success is possible. Where they designate champions and announce targets – like Quaero, the attempt to produce a European Google, or Gaia-X, to produce a European cloud, or the Chips Act which aimed to create a European semiconductor manufacturing sector – all have failed. The list goes on. These are not isolated mishaps. They are symptoms of a technocratic system that confuses announcement with achievement, regulation with capacity, and target-setting with strategy.
The prescription is therefore not another digital-sovereignty programme by the EU Commission. It is them getting out of the way or facilitating their inhibition.
Finally, the report exposes how European political leaders hide behind the digital-sovereignty delusion. National governments remain the locus of democratic authority, taxation, coercion and responsibility. The EU can coordinate, regulate and facilitate, but it cannot substitute for sovereign decision-making. The language of digital sovereignty has become an evasion: a way for national leaders and Brussels technocrats to speak the vocabulary of power while avoiding the burdens that power entails.
Europe’s obstacle is not Silicon Valley or Shenzhen. It is the institutional ised technocratic mindset that prefers performative ambition to the hard, unglamorous work of building capability.
Europe needs chokepoints, not to be choked by the delusion of digital sovereignty.
The future belongs to those who shape it, not to those who regulate or replicate what others have already built.
