European governments are intensifying pressure on Brussels to strengthen the continent’s technological sovereignty as fears grow that powerful new artificial intelligence systems developed in the United States could expose Europe’s critical infrastructure, financial systems, and democratic institutions to unprecedented vulnerabilities.
According to a report published by the Spanish newspaper El País, officials, regulators, and technology experts across the European Union are warning that Europe remains dangerously underprepared for the rise of advanced AI models such as Mythos, the controversial system developed by U.S.-based company Anthropic. The growing concern reflects a broader anxiety that Europe may be losing control over the technological foundations of its own security and economy at a moment when geopolitical tensions between Washington and Brussels are becoming increasingly complex.
The debate has intensified rapidly in recent weeks after the European Central Bank sounded the alarm over the potential risks posed by systems like Mythos, which reportedly possess the capability to identify vulnerabilities in sensitive networks, including banking infrastructure, defense systems, and transportation grids. Officials fear that the accelerating development of these technologies could create severe economic and security disruptions if Europe remains dependent on foreign-controlled AI systems and cloud infrastructure.
Frank Elderson, vice chair of the ECB’s supervisory board, described the situation as urgent. Writing in the institution’s latest newsletter, he warned that Europe must prepare itself structurally for increasingly sophisticated AI systems expected to emerge in the near future. He called for contingency planning capable of addressing what he described as potentially “serious disruptions” triggered by advanced technologies like Mythos.
The warning reflects a growing realization within European institutions that artificial intelligence is no longer simply a matter of innovation or competitiveness. Increasingly, policymakers are treating AI as a strategic issue tied directly to national sovereignty, cybersecurity, and democratic resilience.
That sense of urgency was echoed this week by European Commission Executive Vice President Teresa Ribera during a forum on technological sovereignty at the European Parliament in Brussels. Ribera argued that Europe cannot allow foreign companies or governments to shape the functioning of its economy, services, or political decisions through technological dominance.
“We need to develop our own capabilities,” Ribera said, according to El País. “We cannot allow someone to try to influence our own decisions or the way our economy and services function.”
Yet despite the increasingly strong rhetoric from European officials, critics argue that Brussels has struggled to match its warnings with decisive action. One of the clearest examples involves Anthropic itself, which has reportedly refused to provide European regulators with access to Mythos despite multiple meetings and negotiations with EU officials.
While OpenAI has announced plans to make its newest AI models available to several European companies, including Deutsche Telekom, BBVA, Telefónica, and British cybersecurity firm Sophos, Anthropic has so far limited access to Mythos largely to major American corporations such as Apple, Amazon, Google, Microsoft, Nvidia, and JP Morgan. Those firms are reportedly using the system to identify and patch vulnerabilities within their own networks.
For many European lawmakers, that decision has become symbolic of a deeper imbalance in the global AI race. Dutch Green lawmaker Kim Van Sparrentak criticized Anthropic’s approach, arguing that the company appears willing to trust major American corporations while denying equivalent access to European institutions themselves.
“It is very worrying that Anthropic says it only works with trusted partners,” Van Sparrentak said. “For them, Google is a trusted partner, but apparently the European Commission is not.”
The dispute is unfolding at a particularly sensitive moment for Europe’s AI strategy. Although the European Union was initially praised globally for adopting one of the world’s first comprehensive AI regulatory frameworks, critics now warn that the bloc is weakening its own safeguards under pressure from member states, industry lobbying, and the United States.
In a move that alarmed several technology experts, EU member states and the European Parliament recently agreed to significantly delay the implementation of restrictions on high-risk AI systems. Obligations originally expected to take effect earlier will now only apply from December 2027, a postponement of roughly eighteen months.
The delay has fueled accusations that Europe is retreating from its own regulatory ambitions precisely as the risks surrounding advanced AI systems become more urgent. Critics note that the postponement closely mirrors demands previously made by the Trump administration in communications with Brussels.
Former European lawmaker Ibán García del Blanco, one of the architects of the EU’s AI legislation, warned that Europe risks undermining itself through excessive deregulation disguised as “simplification.” He argued that the continent is weakening its own ability to defend itself technologically at the exact moment stronger safeguards are most needed.
Technology analysts also argue that Europe’s dependence on American cloud infrastructure represents a major strategic vulnerability. Cori Crider, executive director of the Future of Technology Institute, recently published a report warning that Europe’s defense and security systems rely heavily on U.S.-controlled digital infrastructure, potentially exposing critical services to sanctions or remote disruptions ordered by Washington.
Crider warned that the current geopolitical climate makes such concerns impossible to dismiss. She argued that Europe’s strategy of accommodating Washington in hopes of easing tensions is unlikely to succeed, especially under President Donald Trump’s administration.
“Appeasement does not work as a strategy,” Crider said in comments reported by El País. “The only thing that truly protects you from these attacks is resilience.”
Her concerns reflect broader anxieties about the increasingly aggressive posture adopted by the United States toward Europe in technological and economic matters. Several experts involved in the Brussels forum warned that Washington’s dominance in AI infrastructure and cloud computing gives it enormous leverage over European systems.
Some lawmakers went even further, arguing that dependence on foreign-controlled AI tools could leave Europe vulnerable not only economically but politically. Van Sparrentak pointed to recent U.S. national security strategies, which she claims reveal a willingness to pressure European governments more aggressively in pursuit of American geopolitical interests.
“What better way to create chaos than by hacking our transportation systems or our banks?” she warned.
The debate over AI sovereignty is rapidly becoming one of the defining political and economic questions facing the European Union. For decades, Europe relied heavily on regulatory power rather than technological dominance to shape the global digital economy. But the rise of generative AI and increasingly sophisticated cyber systems is exposing the limits of that approach.
Critics argue that regulation alone cannot protect Europe if the infrastructure itself remains controlled by foreign corporations. They say Europe must dramatically increase public investment in domestic technology companies, cloud systems, semiconductor production, and AI research if it hopes to maintain strategic independence.
At the heart of the discussion lies a larger question about democracy, autonomy, and global power in the digital age. Vera Franz, executive director of the Economic Democracy Project at the London School of Economics, warned during the parliamentary forum that Europe faces a stark choice.
“Europe can decide to be bolder and use the influence and resources we have,” she said, “or we can continue signing checks to tech oligarchs and watch our democracy slowly fade away.”
As advanced AI systems become more powerful and more deeply integrated into the world’s economies and security structures, Europe now finds itself confronting a difficult reality: technological dependence may no longer be merely an economic disadvantage, but a direct threat to political sovereignty itself.