Europe is making a bold – if improbable – bet on AI sovereignty. As the continent watches American tech dominance collide with renewed political volatility under Donald Trump, EU policymakers are accelerating efforts to develop competitive large language models entirely within European borders. According to Wired, the push faces steep technical odds but benefits from an unlikely catalyst: mounting concerns about US reliability as a tech partner.

Europe’s relationship with American tech has always been complicated, but it’s reaching a breaking point. The continent that gave us GDPR and the Digital Markets Act is now setting its sights on something far more ambitious: building world-class AI models that can compete with OpenAI, Google, and Meta without depending on Silicon Valley.

The timing isn’t coincidental. As veteran tech journalist Steven Levy reports for Wired, European policymakers see the current US political climate – particularly under Donald Trump’s administration – as both a threat and an opportunity. The unpredictability of American tech policy, combined with growing concerns about data sovereignty and strategic autonomy, has transformed AI independence from a nice-to-have into an existential priority.

But wanting AI sovereignty and achieving it are vastly different things. Europe faces brutal math: developing competitive large language models requires massive compute infrastructure, hundreds of millions in capital, and access to elite AI talent that’s been steadily migrating to American tech hubs for years. Microsoft and Google alone have invested tens of billions into AI infrastructure, while European tech budgets remain fragmented across national borders.

The skepticism is warranted. Building a GPT-4 class model from scratch would require computational resources that dwarf anything currently available in European data centers. Training runs alone can cost upwards of $100 million, and that’s before factoring in the engineering talent needed to optimize these systems. Europe’s track record in consumer tech hasn’t inspired confidence – the continent hasn’t produced a tech giant on par with the American juggernauts since Nokia’s decline.

Yet Europe does have some cards to play. The continent’s strict regulatory framework, once seen as a handicap, could become an asset if companies and governments worldwide grow wary of American tech hegemony. The EU’s proposed AI Act establishes guardrails that might appeal to nations concerned about unregulated AI development. If European models can prove they’re built with privacy and transparency baked in from day one, there’s a potential market among governments and enterprises tired of regulatory whack-a-mole with US providers.

The political catalyst matters too. Trump’s unpredictable approach to international agreements and tech policy has European leaders questioning long-term strategic dependencies. What happens if a future US administration decides to restrict AI model exports or prioritize American interests in ways that conflict with European needs? These aren’t hypothetical concerns anymore – they’re driving budget discussions in Brussels and national capitals across the EU.

France has been the most vocal proponent, with President Macron championing European tech sovereignty for years. Mistral AI, the Paris-based startup, has emerged as something of a flagship for European ambitions, though it’s still orders of magnitude smaller than its American competitors. Germany is pouring resources into AI research centers, while the EU itself is exploring mechanisms to pool funding across member states – a bureaucratic challenge that’s historically proven difficult.

The question isn’t whether Europe should try to build competitive AI – sovereignty concerns are legitimate and the concentration of AI power in a handful of American companies poses real risks. The question is whether the continent can marshal the coordinated effort, sustained funding, and talent retention needed to actually pull it off. Europe’s strength has always been in setting standards and frameworks, not in building the kinds of winner-take-all platforms that dominate consumer tech.

What makes this moment different is the convergence of political will and geopolitical anxiety. If Trump-era policies continue to inject uncertainty into transatlantic tech relations, European resolve might finally match European rhetoric. The continent has the scientific talent, the regulatory sophistication, and now apparently the political motivation. Whether it has the execution speed and risk tolerance to compete in AI’s brutal development cycles remains the multi-billion euro question.

Europe’s AI sovereignty push represents a fundamental recalibration of how democracies think about technological independence. The technical hurdles are immense and the odds of building models that truly rival OpenAI or Google remain long. But geopolitics has a way of reshaping what’s possible when the stakes get high enough. If American unpredictability continues to be a feature rather than a bug of transatlantic relations, don’t be surprised if European determination – and budgets – start defying the skeptics. The real test won’t be whether Europe can match Silicon Valley’s raw compute power, but whether it can build something different enough to matter on its own terms.