ASML Warning Pushes Europe’s AI Sovereignty Debate Up The Stack
ASML chief Christophe Fouquet’s criticism of EU chip-intervention thinking puts the focus on where Europe can still build leverage: trusted compute, data governance, hosting, network interfaces and niche AI supply-chain strengths.

ASML Narrows The Sovereignty Question
ASML chief Christophe Fouquet’s warning over Europe’s chip-intervention approach puts a harder question in front of policymakers: which parts of the AI supply chain can Europe realistically influence.
The argument is not that semiconductors no longer matter.
It is that the AI stack now links chips, memory, cloud infrastructure, packaging, telecom networks, power systems and capital markets in ways that no single region fully controls.
The policy target is the EU’s evolving Chips Act 2.0 framework, which is moving from subsidy-led capacity building toward supply-chain coordination in shortage conditions.
Fouquet’s concern is that directing chip flows from the top down may overstate Europe’s leverage at the base layer.
ASML is a global choke point in advanced lithography, but the EU accounts for only one percent of EUV sales, limiting how far regional political control can reach through one supplier.
That limitation matters because semiconductors sit inside many strategic systems at once.
The same supply discussion touches data centres, power grids, telecom networks and defence systems.
A rule aimed at one shortage can therefore collide with memory, packaging and cloud dependencies that are controlled elsewhere and priced through global markets.
Control Moves Toward Compute And Trust
The practical alternative is not to abandon sovereignty.
It is to move the leverage point higher.
France and the UK are described as taking a more pragmatic route around sovereign AI by focusing on compute infrastructure, data governance, trusted hosting, niche specialties and national assets.
France brings nuclear power into the discussion, while the UK points to supercomputers, chip design, GPU networking, R&D clusters and early-stage startups.
That approach fits a layered AI economy better than an attempt to command every supply input.
Critical sectors still depend on memory, cloud, packaging and other supplies controlled outside Europe.
The more workable question is where jurisdiction, access rules and trusted infrastructure can create durable influence for governments, telcos and industrial users.
Space Networks Do Not Remove Jurisdiction
The same logic applies beyond terrestrial infrastructure.
SpaceX is presented as a parallel infrastructure layer that combines rockets and satellites, fills rural coverage gaps and could eventually support niche segmented compute workloads.
But orbital systems do not erase geography.
They still rely on ground stations, spectrum authorisations, regulatory frameworks, gateways, user devices and Earthbound supply chains.
The constraints are practical as well as legal.
The source frames space infrastructure as a way around some planning and connectivity bottlenecks, not as a full cloud layer above national control.
Latency, transport, maintenance and economics remain part of the infrastructure equation, which keeps terrestrial policy and network interfaces relevant.
That makes sovereignty more distributed, not less relevant.
Control points now sit across terrestrial, subsea, airborne and orbital systems.
Europe’s next test is whether its AI policy can identify those interfaces instead of treating the whole supply chain as a hierarchy that can be commanded through chip allocation.
The concrete watchpoint is how the EU’s Chips Act 2.0 process balances shortage coordination with the higher-stack controls around compute, hosting, data and network access that France and the UK are already emphasising.















