Europe Wants AI Sovereignty: Macron, Trump and Claude

European leaders, startups and researchers are pressing for AI sovereignty after the United States moved to restrict access to Anthropic’s Claude Fable, a step that prompted Anthropic to pull the model off the market and prompted fresh political urgency in Europe.

What triggered Europe’s renewed push for AI sovereignty?

Europe’s push intensified when the Trump administration attempted to place Anthropic’s Claude Fable under strict export control regulations, denying foreigners access and prompting Anthropic to pull the model from the market. The move signaled to European policymakers and companies that they cannot rely on American models if US policy can cut access at any time, and it came as Vivatech in Paris and the G7 in Evian-les-Bains focused attention on the issue.

That regulatory action followed a landscape where US and Chinese firms dominate the frontier. The article cites Anthropic’s recent $65 billion fund-raise as a point of contrast with the sums invested in European and UK AI startups last year, and it notes declining European enrollment in US universities as part of a broader talent shift.

How are governments and companies responding?

European governments and industry are coupling public pledges, cross-border partnerships and startup momentum: France’s "Choose France" initiative has won pledges of over 100 billion euros in AI infrastructure, anchored by Softbank’s 75 billion-euro commitment to build data centers in France, pending approvals. Private efforts include multinational partnerships such as Cohere’s collaboration with German firm Aleph Alpha and a memorandum of understanding with Spain’s Indra.

Cohere CEO Aidan Gomez told attendees that democracies must occupy the number two position in global AI supply, saying, "We need to ensure that a democracy occupies the number two position, and that's not true today." Yann LeCun, who recently resigned as Meta’s chief AI scientist, is pursuing Project Tapestry, an effort to build an open foundation model that governments and industry could use to create region-specific assistants. CEOs of smaller European firms say the US action has already changed conversations: Michael Förtsch of Qant said the short-lived export controls on Fable "triggered a complete new discussion on sovereignty in Europe."

Jakob Uszkoreit, CEO of biotech firm Inceptive, described a potential talent shift away from US labs, saying he could assemble an all-star European team if they were offered "reasonable personal incentives" and the ability to do their best work. The article also notes that seven of the eight coauthors of the Transformers paper were foreign-born, underscoring how much frontier AI talent is international.

Why it matters

The export-control episode reframes sovereignty as survival: a foreign company’s business built on US models is vulnerable if those models become inaccessible. European pledges of over 100 billion euros and large private commitments aim to buy capacity and time, but the piece stresses that matching US ecosystems will require unprecedented cross-border cooperation, a shift from risk aversion to a moonshot mentality and long-term investment.

Europe’s strategy mixes public infrastructure pledges, partnerships to pool engineering and data-center resources, and efforts to attract talent back from US labs. If those elements do not materialize, European firms risk remaining dependent on American models that can be restricted by US policy.

What to watch

Watch whether the Choose France infrastructure commitments are approved and executed, and whether Project Tapestry or multinational partnerships such as Cohere plus Aleph Alpha produce a widely adopted, open foundation model. A concrete signal of success would be new European-trained frontier models or services that remain available to European customers even if US export rules tighten further.