Trump’s return highlights why Europe must build its own AI ecosystem—secure, competitive, and true to its values.
The succession of global shocks over the past two decades – including the 2008 global financial crisis, the COVID-19 pandemic, Russia’s invasions of Ukraine, and rising inflation – has fuelled a resurgence of industrial policy across Europe. But the added shock of Donald Trump’s return to the White House has underscored the urgency of bolstering domestic economic resilience and integrating national and regional security priorities into economic policymaking.
For the European Union and the United Kingdom – which are now facing an openly hostile US administration – the digital economy must become a central focus. Former European Central Bank President Mario Draghi’s report on EU competitiveness and the UK’s AI Opportunities Action Plan both highlight digital and AI technologies as potential engines of future innovation and growth.
The challenge, however, lies in reducing Europe’s heavy reliance on major US tech firms. Whereas President Joe Biden’s administration took a confrontational approach toward Big Tech, the Trump administration has signalled that any future trade agreement will hinge on the EU scaling back its push for new digital regulations and taxes.
For Europe, business as usual is no longer an option. Policymakers must develop a coherent and strategic alternative to reliance on American technology. If this sounds like a quixotic endeavor, consider the creation of Airbus, which started as Europe’s answer to Boeing. An “Airbus for AI” – a publicly funded, commercially operated alternative to US-based platforms – is both feasible and necessary.
Recent experience has exposed the vulnerability of Europe’s public services and private sector to the whims of US tech executives whose top priority is to remain on good terms with their own government. Elon Musk’s willingness to breach Starlink contracts with European governments has fuelled concerns about the reliability of American platforms, as have efforts by other US firms to exploit trade tensions to lobby against European tech regulations like the EU’s Digital Markets Act and the UK’s Online Safety Act.
Consequently, even if the Trump administration were to reverse its protectionist policies tomorrow, doubts about the trustworthiness of US tech companies would persist. And since Chinese tech platforms are equally problematic, it is increasingly evident that European governments must begin developing an independent digital and AI ecosystem.
Politicians often shy away from necessary decisions that seem too difficult or costly. Yet in a new policy brief, Joshua Tan, Brandon Jackson, and I argue that a public-private commercial model for building large-scale European AI systems – known as foundation models – is technically and financially viable, and could be launched quickly.
While European research laboratories and institutes already collaborate, they still lack a coordinated product strategy and a clear pathway from innovation to market. Airbus should serve as a blueprint. The company was established in 1970, when European governments recognized that their national aviation firms were too fragmented to compete with Boeing. France and Germany – later joined by the UK and Spain – formed a consortium that pooled expertise, resources, and funding to develop cutting-edge, innovative aircraft.
Although the Airbus project emphasized commercial viability from the outset, it was also supported by public financing and industrial-policy tools such as advance purchase commitments, research investment, and technical training. Domestic political imperatives were addressed by distributing the benefits among participating countries through supply-chain specialization.
The Airbus project achieved a major breakthrough with the A300, demonstrating the potential of aligning scientific research with market needs. Developing a twenty-first-century AI equivalent will require similarly large infrastructure investments and sustained political commitment.
Fortunately, the building blocks for such a model are already in place and could be mobilized quickly by a select group of European governments. Many are already investing in national public computing facilities, and several leading European labs – such as those participating in LLMs4Europe – have begun releasing cutting-edge foundation models that push the boundaries of AI research.
Coordinating these efforts will be essential to achieving the necessary scale. In addition to creating value for taxpayers, this will help European alternatives differentiate themselves from dominant US tech firms – a key condition for commercial viability. Private companies like France’s Mistral show that a robust European market could be catalyzed and given a defensive moat through coordinated strategic action. Singapore has also demonstrated the potential for effective public-private partnerships – though not yet at scale – and other successful models have emerged in the semiconductor sectors of Japan and the US.
Given US incumbents’ first-mover advantage, any European initiative will require significant public support. But such support doesn’t have to come solely through direct funding. It can also take the form of access to public computing resources, tax advantages tied to public-interest commitments, preferential access to government datasets unavailable to US firms, and purchasing commitments from public institutions.
Moreover, market demand can be further stimulated through targeted measures. For example, the EU’s AI Champions initiative, which focuses on high-value industrial applications, could serve as a powerful catalyst once a European consortium is ready to supply the necessary technology.
The strategic and economic case for developing an alternative to US and Chinese AI is compelling. But a public-private, commercially driven European AI initiative should do more than serve industrial policy and security objectives; it must also reflect the culture and values of the continent’s middle powers. American-developed AI models increasingly project a US-centric worldview – one that is rapidly, and perhaps irrevocably, diverging from Europe’s. It is time for European countries to chart their own course.
Diane Coyle is professor of public policy at the University of Cambridge. She is the author, most recently, of Cogs and Monsters: What Economics Is, and What It Should Be (Princeton University Press, 2021).