While Washington tightens controls on the export of advanced artificial intelligence technologies, Brussels is making an unexpected move. Austria has taken the initiative to propose that Anthropic — one of the world's leaders in AI — relocate to the European Union. This proposal, announced just a few weeks after the United States imposed restrictions on foreign specialists' access to the most powerful AI models, appears to be an attempt by the EU to seize the strategic initiative.
Strategic challenge or a gesture of desperation?
EU Digital Secretary Alexander Prell has sent a corresponding appeal to the European Commission. In his letter, he emphasized that Europe risks being left on the sidelines of key discoveries in the field of artificial intelligence if it does not take decisive action now. The essence of the proposal is to "strategically place Anthropic and ensure its participation in the life of the European Union." In return, the company is offered legal guarantees, fresh investments, and full access to the single European market.
However, as Prell himself notes, specific amounts, timelines, and a clear implementation plan are currently lacking. Moreover, many within the EU itself are skeptical of this idea. And this is not surprising — the proposal looks more like a gesture than a well-developed strategy.
The reason — US restrictions and data "leakage"
The catalyst for the European initiative was the actions of the US Department of Commerce. On June 12, a directive was introduced restricting the export of Anthropic's latest models — Claude Fable 5 and Claude Mythos 5. The reason was the discovery of instructions for cyberattacks in the Mythos model, which caused concern in Washington. Interestingly, the trigger was an appeal from Amazon, Anthropic's largest investor, which has poured $13 billion into the company. As a result, both products temporarily became unavailable to all users, as the company cannot filter them by citizenship.
On June 26, the US partially eased export restrictions for more than 100 trusted American organizations, but the Fable 5 model remains under ban.
Why Europe will find it difficult to lure the giant
Despite ambitious statements, the EU has fundamental problems. Anthropic is deeply rooted in the American ecosystem: the company is investing $50 billion in building data centers in Texas and New York, and its main training partner is Amazon, which has already pledged to spend over $100 billion on cloud services over a decade. Additionally, Anthropic estimates the AI needs in the US by 2028 at 50 gigawatts of new energy capacity — a colossal volume that Europe simply does not have.
The European Chips Act, aimed at achieving 20% of the global chip market by 2030, according to independent auditor estimates, is unlikely to exceed 11.7%. There is a shortage not only of computing resources but also of venture capital and a developed energy base. It is these factors that keep businesses in America.
My expert opinion: The EU's attempt to lure Anthropic is more of a political gesture than a realistic business strategy. Europe is strong in regulation but weak in creating conditions for scaling capital-intensive AI projects. Until Brussels resolves issues of energy and access to cheap capital, such initiatives will remain mere good intentions. The AI market will consolidate where resources exist — in the US and, possibly, in Asia.