Last weekend proved a simple truth.
Europe does not control the technologies it increasingly depends upon.
The Trump administration’s decision to restrict overseas access to Anthropic’s latest Claude Fable 5 and Claude Mythos 5 models should serve as an urgent wake-up call for Europe.
Whether one agrees with the US move or not is almost beside the point.
The bigger lesson is clear. If critical technologies are controlled elsewhere, access to them can ultimately be controlled elsewhere too – or have the plug pulled without warning.
At a time of major global instability and uncertainty, semiconductor export controls, growing US-China competition, off-the-scale Russian interference and an increasingly uncertain geopolitical environment, technological dependence has become a strategic issue.
At DesignVerse, we spend our time building AI systems for highly regulated mission-critical industries – from defence and financial institutions to air traffic control – where reliability, security and resilience are non-negotiable.
These sectors are not interested in technology trends or headlines. They are focused on top security operational resilience.
That is why the debate around technological sovereignty is no longer academic. It’s urgent.
Europe’s dependence has become a vulnerability
For decades, Europe outsourced much of its technological future. That approach made sense in a more predictable world. It no longer does.
Artificial intelligence, cloud infrastructure and software platforms increasingly underpin everything from defence and energy to healthcare, aviation and public services.
These technologies are no longer conveniences. They are strategic infrastructure.
Yet Europe remains heavily dependent on non-European providers. More than 70% of Europe’s cloud infrastructure is estimated to rely on three US hyperscalers.
Even data stored within Europe by US cloud providers remains subject to American legislation, including the CLOUD (Clarifying Lawful Overseas Use of Data) Act and FISA (Foreign Intelligence Surveillance Act), meaning localisation alone does not necessarily guarantee sovereignty.
No nation would willingly place its energy grid or defence capabilities under the full control of another country.
Yet in digital infrastructure, Europe has become dangerously exposed.That dependence is no longer merely an economic issue, it is becoming a geopolitical vulnerability.
The Mythos-5 saga exposed a reality Europe can no longer ignore
While many of us in the European technology ecosystem have warned about these risks for more than a year to no avail, this weekend’s restrictions should be viewed as a DEFCON 1 watershed moment in Brussels.
Not because America acted wrongly. Every nation has the right to protect technologies it considers strategically important. But the decision revealed something more profound.
The world’s most advanced AI systems are no longer being treated as products. They are being treated as strategic assets.
American officials justified the restrictions partly because of concerns over the capabilities of these increasingly powerful systems and the risks they could pose in the wrong hands.
That should give Europe pause.
If these technologies are considered sufficiently powerful to warrant export controls, should Europe really be comfortable relying almost entirely on systems it neither controls nor governs?
Cybersecurity leaders have spent decades warning against single points of failure.
Yet in AI, we risk creating precisely that. Concentrating advanced models, data platforms and organisational knowledge in the hands of a small number of companies creates both geopolitical and cyber risks.
Strategic dependence is not the same thing as partnership. Partnerships are strongest when both sides possess meaningful capability.
Europe’s defence industry already understands the stakes
Over recent months, DesignVerse has been in discussions with some of Europe’s largest defence contractors.
They approached us specifically because of our ability to use AI to build secure, mission-critical software at speed using European-grown technology tailored to their clients’ needs.
One message emerged consistently. Strategic autonomy in technology is a matter of national security.
Executives and cybersecurity leaders recognise that Europe cannot build sovereign defence capabilities while remaining heavily dependent on technologies and AI systems controlled elsewhere.
The issue is not distrust. It is resilience. True sovereignty means retaining the ability to build, maintain and control the systems upon which societies depend.
Europe’s defence sector understands this instinctively. The rest of the economy needs to reach the same conclusion.
The Palantir paradox
Perhaps nowhere is this debate more visible than around data.
Palantir has built impressive capabilities and won major contracts across Europe, including the UK Ministry of Defence, NATO’s Maven programme and projects involving intelligence and law enforcement agencies across various EU states.
Its success reflects technological prowess. But it also raises important questions.
Can Europe genuinely speak about digital sovereignty while some of its most sensitive data and decision-support platforms become increasingly concentrated within a handful of non-European providers?
This is not an argument against Palantir or against American companies. It is an argument against concentration. No single company, country or platform should become indispensable.
Because dependency and sovereignty are ultimately incompatible.
Regulation alone will not solve the problem
Europe has become exceptionally good at regulating technology. But regulation does not create technological leadership. Innovation does. Execution does. Building does.
For too long, Europe has celebrated strategy papers while outsourcing execution.
The continent already possesses extraordinary engineering talent. Romania and the wider Central and Eastern European region are proving this every day, as a recent event in the European Parliament highlighted clearly.
At DesignVerse, we also recently helped modernise mission-critical aviation software used across Europe’s air traffic systems in partnership with EUROCONTROL, rebuilding a fifteen-year-old application in just over a month.
Europe does not lack talent. It lacks urgency, political support and financial backing.
What Europe should do next
Technological sovereignty does not mean isolation. Nor does it mean replacing Silicon Valley.
But Europe needs a clear plan. That means investing in sovereign digital infrastructure.
It means supporting defence and AI integration programmes. It means providing scale-up funding capable of retaining Europe’s best startups and engineering talent.
It means creating stronger incentives for European companies to build and deploy strategic technologies at home. And it means recognising that sovereignty is not about protectionism.
Operationally, sovereignty means retaining the ability to choose, control and evolve the technologies upon which our economies and societies depend.
Because in the age of artificial intelligence, sovereignty is no longer optional. It is essential.
