BRUSSELS – A new narrative is gaining traction across Europe: the continent can rebuild its technological power by leveraging its longstanding advantages in mechanics, automation, and engineering. Combined with its deep-rooted open-source tradition and next-generation artificial intelligence (AI )models, these assets could make Europe an industrial AI superpower. While this vision is promising, can it really reverse Europe’s slide toward technological decline?

Such a strategy calls for the development of a distinct AI paradigm centered on smaller, specialized models, distributed approaches, open-source collaboration, and high-quality industrial data. A recent report estimates that Germany alone could unlock 1.7 trillion euroes ($2 trillion) in untapped high-tech potential by 2030.

It is encouraging to see Europe’s pessimism about its place in the global AI race give way to renewed confidence. Drawing on its domain expertise, world-class university research, and open-source tools that enable interoperability and collaboration, the European Union hopes to become a global leader in robotics, advanced machinery, aerospace, and defense technologies, helping to shape an AI future in which data quality and specialization matter more than sheer scale.

European politicians love this narrative, which promises a technological resurgence without confrontation. By building on existing strengths, the EU can ostensibly avoid redirecting demand away from foreign providers and sidestep sovereignty debates that are putting it at odds with U.S. companies and U.S. President Donald Trump’s administration.

The results are already visible. Following intense lobbying by U.S. tech firms – often channeled through American officials – political leaders have begun distancing themselves from France’s proposal to introduce a “European preference” requirement into the bloc’s industrial-policy toolkit. The proposal was tied to the Industry Acceleration Act (IAA), promoted by the EU’s Prosperity and Industrial Strategy Executive Vice-President Stéphane Séjourné and supported by French President Emmanuel Macron.

US tech giants, which dominate Europe’s digital infrastructure, have aligned with European industry groups to oppose the IAA, pushing back against digital sovereignty which several governments had previously committed to support. Their efforts, evident in the rhetorical shift from “Made in Europe” to “Made with Europe,” include selective invocations of World Trade Organization rules and a pivot toward a “different” tech growth model that would ultimately leave the status quo intact.

While this path may appear promising, Europe cannot wish away its technological dependencies with talk of “new model architectures.” From cloud services and connectivity to the digital backbone – browsers, search engines, office software, and operating systems – the EU’s reliance on non-European infrastructure runs deep.

That dependence carries real costs: reduced resilience and security, lost revenue for European developers, underinvestment in domestic assets, and diminished capacity to shape innovation. Unless these structural weaknesses are addressed, Europe’s economic prospects will remain bound to others’ infrastructure.

Europe cannot build its AI future through magical thinking. Industrial ambitions – even those backed by domain expertise, valuable data, and open-source tools – will not change Europe’s status as a digital colony of a handful of mostly U.S. firms. In today’s geopolitical landscape, only ownership and control of infrastructure can guarantee strategic autonomy. Europe’s drive for technological sovereignty, including initiatives like EuroStack, seeks to address precisely that structural constraint on leadership in industrial AI.

It is hardly surprising that Europe’s effort to claw its way to technological relevance has encountered intense resistance, as the U.S. tech giants that have blanketed the continent with their infrastructure are not inclined to make room for local competitors. This may well be the coming decade’s defining economic battle, with the outcome decided by whether Europe can direct its demand toward domestic developers and suppliers.

The pushback from the U.S. giants often takes a subtle form, with sovereignty presented as an add-on to infrastructure generously provided by American hyperscalers. Nothing illustrates this better than the growing number of events on tech sovereignty hosted by the very companies whose dominance is at issue. Their message to Europe is unmistakable: build whatever you want, but use our AI construction kit.

At the same time, orthodox free-trade narratives are increasingly being weaponized to pressure European governments that pride themselves on economic openness. Any attempt to direct European demand toward domestic suppliers, the argument goes, is incompatible with the rule of law and established WTO practices.

Resisting these narratives will not be easy. Over the past two decades, hyperscalers have embedded themselves within Europe’s institutional fabric, building extensive lobbying operations, establishing revenue-sharing partnerships with system integrators, and extending their reach into universities, research programs, advisory boards, and standard-setting bodies.

The resulting familiarity has fostered a kind of learned helplessness that continues to shape risk assessments across the public and private sectors. Procurement frameworks favor incumbent architectures, entrenching legacy systems as the default and casting alternatives as risky and uncertain, even when they are technically viable. Likewise, calls to recalibrate demand are resisted because they challenge established economic relationships, institutional habits, and longstanding influence networks.

European policymakers must reject efforts to reframe sovereignty as a trade issue. The core challenge is structural: demand, institutions, and ecosystems have been shaped in ways that reinforce dependence, and now they need to be redesigned.

Given that industries scale where there is sustained, addressable demand, building productive capacity requires demand-side tools, including local-preference mechanisms. To paraphrase U.S. Supreme Court Justice Robert Jackson, trade rules are not a suicide pact. Every major economy balances compliance with strategic domestic priorities, and the EU should do the same.

This is especially true for digital infrastructure. Far from neutral commercial inputs, these are critical systems woven through public administration, defense, and economic life. Resilience and long-term capacity must take precedence over efficiency.

Above all, European builders must reclaim a meaningful role in their own market. Political leaders cannot speak of sovereignty and then retreat at the first sign of lobbyist pressure. Forging an AI-driven future requires building durable productive capacity and channeling demand toward domestic suppliers. At a time when major powers are pursuing aggressive industrial policies, the EU cannot afford to remain on the sidelines.

Cristina Caffarra is chair of the EuroStack Industry Foundation and co-founder and vice chair of the Competition Research Policy Network at the Centre for Economic Policy Research. This article was distributed by Project Syndicate.

© The Korea Times