Earlier this month I attended Vivatech, a huge tech conference in Paris. One fear dominated the discussions: the prospect of ending up stuck using American AI, trained on American values. While the US and China are locked in an AI arms race, France and Germany, which consider their engineering talent second to none, feel boxed out. Not only are they demanding to be heard, but they are touting plans to address the situation. If “sovereignty” was your word in a drinking game, you’d be pickled within three hours.
In my decades of reporting on tech, I’ve covered multiple efforts by countries to replicate the Silicon Valley effect. While there have been plenty of individual success stories, no country or market has come close to matching the ecosystem and mindset that gave rise to companies such as Google, OpenAI, and Anthropic. While investors throw boatloads of cash at American companies, Europeans get relative crumbs. One statistic I heard several times last week was that Anthropic’s recent $65 billion fund-raise was more than the entire sum invested in European and UK AI startups last year. Actual results reported by the EU seem to bear this out.
Nonetheless, sovereignty discussions at Vivatech were infused with hope. Optimists cited significant new funding, collaborative efforts, and next-generation technology that might not be as resource-intensive as the leading large language models. And several cited a wild card that might be the biggest boon to European tech in decades: Donald Trump.
Vivatech overlapped with the G7 conference in Evian-les-Bains, France, where French president Emmanuel Macron lectured AI executives on the sovereignty issue. If the US continued down its path of nationalistic AI, he said, France would take steps to go it on its own. Aiden Gomez, the CEO of Toronto-based Cohere, also tried to convey his sense of urgency to the crowd in Evian. “We need to ensure that a democracy occupies the number two position, and that's not true today,” Gomez told me at Vivatech. “I think the G7 understands that we need a diverse supply chain of AI providers.”
It sounds almost delusional for Europe to think that it can build the world’s second-best AI. More than 20 nations would need to work closely together, overcome their continental impulses to strangle innovation with red tape, and lure unprecedented sums of investment. Most of all, Europe must shift from a risk-adverse mindset to a moonshot mentality. But Macron has made some progress. His “Choose France” initiative has won pledges of over 100 billion euros in AI infrastructure, anchored by Softbank's 75 billion-euro commitment to build huge data centers in France—pending approvals, of course.
As for collaborations, Gomez tells me that Cohere is trying to stitch together a multinational chain of partnerships, beginning with one with the German AI firm Aleph Alpha. The idea is to pool resources in both engineering and infrastructure for a “sovereign-first” approach. “A few weeks ago, I was with the king of Spain to sign an MOU with Indra, which is the largest tech company in Spain,” he says.
Yann LeCun, the AI pioneer who recently resigned as Meta’s chief AI scientist, is pursuing Project Tapestry, a massive effort among governments and private industry to join forces in building a state-of-the-art frontier foundation model. “The governments of the world all want AI sovereignty,” he says. “The only way I can see this happening is if there’s an open, free foundation model, on top of which anybody can build their own specialized assistant for their own language, culture, value system, and political biases.”
While all these plans appear reasonable, they seem all too similar to previous failed efforts to elevate European tech. But in 2026, there’s a new element: the Trump administration's clumsy policies that make the status quo utterly intolerable. For generations, the US has lured Europe’s best scientists; they now feel unwelcome. European enrollment in US universities is down. Jakob Uszkoreit, CEO of the AI-based biotech firm Inceptive, says that even by the end of Trump’s first term, he saw talent moving away from the US, a trend that has gained more steam in the current administration.
It wouldn’t take much to significantly drain that talent from US shores, Uszkoreit says. “I’d have no problem assembling an all-star team of Europeans, many of which would leave their very cushy current US frontier lab jobs, provided two things: reasonable personal incentives and they have to be able to do their best work,” he says. (By the way, both Uszkoreit and Gomez are prime examples of what the US might miss if foreign researchers don’t come to America—they are among the eight coauthors of the famous Transformers paper that spurred generative AI. Seven of the eight were foreign-born.)
Earlier this month, the Trump administration gave European tech perhaps its biggest incentive yet to supercharge its own ecosystem. That was when it attempted to limit Anthropic's powerful Claude Fable model by putting it under strict export control regulations, denying foreigners access. Argue all you want over whether this clampdown was justified—from the European point of view it was a wake-up call that sovereignty now equates to survival. A foreign company can’t build a reliable business around Claude or any other American model if at any point the US government can deny access. Even more dire, the designation bans access even by foreign nationals who work at Anthropic, even those who helped develop Fable! Anthropic quickly pulled the model off the market, but the message had been sent: Europeans can’t count on American companies.
European startups are already getting a salutary push from the US government’s actions. “The attention we currently have in Europe would not have been the same without Trump,” says Michael Förtsch, CEO of Qant, a chip startup. The short-lived export controls on Fable, he says, “triggered a complete new discussion on sovereignty in Europe.”
“Europe had gotten pretty complacent in light of a very reliable, well-aligned, near-hegemonic [situation],” says Uszkoreit. “The US just made it clear that in the new world order, that’s over.” Europe has no choice but to pursue AI sovereignty.
This is an edition of Steven Levy’s Backchannel newsletter. Read previous newsletters here.
Instagram Engineering
Shopify Engineering