Austria has asked the EU to explore hosting Anthropic inside Europe, weeks after Washington restricted foreign access to the company’s most advanced artificial intelligence models.

State Secretary for Digitalization Alexander Pröll made the proposal in a letter to the European Commission. He admitted he could not say how the plan would work in practice.

Europe’s Bid to Host Anthropic

Pröll sent the letter to European Commission Executive Vice President Henna Virkkunen. He argued that Europe risks being cut off from frontier AI breakthroughs unless it acts now.

In his letter, Pröll asked member states to weigh a far bigger step.

“the strategic establishment and participation of Anthropic within the European Union”

His pitch dangled incentives such as legal certainty, fresh capital, and full access to Europe’s single market. He offered no funding figure, timeline, or build plan, and conceded skeptics would doubt whether the idea can work.

Brussels is already weighing fallout from the public controversy.

Follow us on X to get the latest news as it happens

The US Curbs That Triggered the Appeal

Washington set this in motion on June 12. The Commerce Department issued an export directive on the firm’s two strongest models. It barred every foreign national, even Anthropic’s own non-citizen staff.

The order hit Claude Fable 5 and Claude Mythos 5, its newest AI models, days after launch. Unable to screen users by nationality, Anthropic pulled both worldwide. Claude Opus 4.8 stayed online.

Officials cited national security. The warning came from Amazon, Anthropic’s biggest backer, after its researchers pulled restricted cyberattack guidance from Mythos. Anthropic CEO Dario Amodei called the bypass narrow, not a full jailbreak. The model had already shown it could crack guarded government systems.

On June 26, the government eased the export block for more than 100 trusted US institutions. Fable 5 stays restricted.

Any relocation runs into Anthropic’s American foundations. The firm is funding a $50 billion data center build in Texas and New York. Amazon has invested $13 billion and is its primary training partner. In return, Anthropic has committed to spend over $100 billion on Amazon’s cloud within a decade.

The company also estimates US AI will need roughly 50 gigawatts of new power by 2028. Europe sits far behind on the inputs. Its Chips Act targets a 20% share of global chip output by 2030, up from under 10% now. The bloc’s own forecast sees just 11.7%, and EU auditors call the goal very unlikely.

The appeal lands as Brussels weighs its answer to US control over frontier AI. Europe still leads on regulation through its own AI rules.

Yet it lacks the compute, capital, and power base that keeps Anthropic rooted in America. Whether the Commission sees a real option or a political signal should grow clearer soon.

Источник: BeInCrypto