Europe has a chance to take control of its AI future
Europe has the policy framework, the investment capital, and the political will, writes tech expert Bjorn Hovland.
European dependency on foreign technology has been in the news recently with breathless headlines about ‘kill switches’ and risks to critical national systems from reliance on US and Chinese technology.
Listen to this article
On June 12, the White House restricted access to Anthropic's Fable 5 and Mythos 5 for all non-US users without warning. Anthropic has since announced plans to restore access, but the episode cut through years of policy debate and made the stakes concrete.
Critical infrastructure dependency can be switched off at any moment by a decision taken in Washington, with no notice and no technical recourse for the organizations affected.
Consider what that means in practice. Hospitals, banks, and national defense systems running AI that goes dark in a crisis, a conflict, or a diplomatic standoff. Customers unable to access deposits. Airlines grounded. Military operations halted.
These are nightmare scenarios which Europe’s leaders are now confronting, not least because EU countries rely on non-EU tech for 80% of products and services. The next Fable 5 situation will not be the last.
The European Commission recognised the urgency. Its European Technological Sovereignty Package, released June 3, includes plans to triple data center capacity, steps to strengthen the European semiconductor industry, an “open source-first approach”, and measures to make data centers more sustainable given their substantial energy demands.
At the national level, the UK announced a £500 million Sovereign AI Fund and a £1.1 billion AI Hardware Plan, while France committed €109 billion to AI infrastructure last year. The political will is real and the capital commitments are serious.
But the central question deserves a direct answer:can these ambitions translate into genuine sovereignty? Yes, if three things happen.
Some analysts have described this as a €4 trillion problem, without any guarantee of success. If countries are banned from using leading foreign tech, will users have to settle with third-rate products, buggy systems, and systems that are ultimately expensive and uncompetitive?
The good news for Europe is this is a genuine moment of opportunity.
Europe can not only delink from US and build her own independent capabilities, but by boosting domestic capacity and skills, there could be huge economic, strategic, and employment advantage.
Shift the policy agenda from regulation to innovation.
I don't believe there is a European alive who hasn't experienced the frustration of GDPR firsthand, a regulation designed to protect citizens that in practice has made sensible exchanges of information harder while doing little to reduce dependence on the foreign platforms it was meant to constrain.
Mario Draghi's landmark 2024 report got this right: innovators need support, competitive tax environments, and far less bureaucratic burden. That prescription is more urgent today than when he wrote it. Europe will not innovate its way to sovereignty through compliance frameworks.
Make Europe the most attractive place in the world to build AI infrastructure.
Data centres that take years to permit should take months. Investors committing capital to European infrastructure need confidence they will not be sacrificed on the altar of political expediency. The capital is ready to move. The question is whether Europe captures it or watches it deploy elsewhere. A continent serious about sovereignty builds the conditions that make it the destination of choice, not a harder place to operate than the alternatives.
Commit to open source as the sovereign infrastructure foundation.
This is where the structural answer to the kill switch problem lives. Once open source software is deployed on infrastructure a European institution owns and operates, no foreign government can shut it down, restrict access to it, or reach into the system remotely.
Open source enables local fine-tuning and management of sovereign data without dependence on the proprietary systems of companies like OpenAI or Anthropic. Because the code is open and auditable, it provides a level of security and trust that closed foreign technology cannot match.
The open source model makes the national origin of a technology provider structurally irrelevant. What determines sovereignty is whether the infrastructure runs on hardware you own, code you can read, and systems you control end to end.
European institutions should require, as a condition of any AI infrastructure deployment, that the software runs on European-owned, open source infrastructure. That single procurement condition is what converts any vendor relationship into a sovereign one. The CADA Level 4 framework the Commission has already established provides the legal basis to demand exactly this.
There is a further advantage that belongs specifically to Europe: energy.
Europe's net zero commitments and the structural cost of energy make AI infrastructure efficiency a competitive priority in a way that does not apply in the same way elsewhere.
Open source infrastructure compounds efficiency gains at every layer of the stack. The organizations that optimize for this will produce more from every watt of energy consumed and every euro of capital deployed. Europe should be leading on this metric, not catching up to it.
If Europe moves with genuine urgency, the prize is significant: a frontrunner position in the multi-trillion euro global AI market, high-value employment across the full stack from chip design through model development and applied enterprise AI, and the strategic autonomy to operate on its own terms in an increasingly uncertain world.
The alternative is to follow the path of Ming China after it abandoned maritime exploration in 1433, a civilization that held every advantage and chose not to act on it.
Europe has the policy framework, the investment capital, and the political will. What is required now is the speed and conviction to act on all three before the next Fable 5 makes the decision for her.
____________________
Bjorn Hovland is the President of CIQ, the enterprise infrastructure software company.
LBC Opinion provides a platform for diverse opinions on current affairs and matters of public interest.
The views expressed are those of the authors and do not necessarily reflect the official LBC position.
To contact us email opinion@lbc.co.uk