Austria's pitch to host Anthropic inside the EU lays bare a truth Brussels does not like to admit: Europe has no frontier AI model it controls, no guarantee it can keep using the ones America builds, and no clear plan to fix either.
State Secretary for Digitalisation Alexander Proell sent a letter to EU tech sovereignty chief Henna Virkkunen urging the bloc to explore "the strategic establishment and participation of Anthropic within the European Union." The trigger: a June 12 Commerce Department directive forcing the company to cut foreign access to its two most advanced systems, Mythos 5 and Fable 5.
Anthropic could not implement a nationality-based restriction on its shared cloud infrastructure, so it shut off both models for everyone outside the US. Europe went dark on two of the most capable AI systems in existence, not because of anything it did, but because of a Washington decision it had no say in.
"Let us jointly explore the strategic establishment and participation of Anthropic within the European Union, with legal certainty, market access, capital and a set of values that suits this company," Proell wrote. The letter does not specify how this would work, no European subsidiary, no data-residency framework, no equity stake.
Proell acknowledged there would be skepticism about feasibility. "The real question is not whether it is easy," he wrote. "The question is whether we Europeans are prepared to be the architects of our technological future, or whether we wish to remain mere administrators of decisions made elsewhere."
The timing is awkward. On Friday, Commerce Secretary Howard Lutnick partially lifted the export ban, allowing roughly 100 companies and federal agencies to access Mythos 5 again. Fable 5, the more capable of the two blocked models, remains blocked with no timeline for reinstatement.
The partial reversal undercuts the urgency of Austria's pitch but does not erase the structural problem it identified. The company has not commented on the proposal. The US IPO it is reportedly preparing for would likely be blocked by the US government if it tried to relocate, and leaving would cost it access to its largest market.
The deeper issue is one he left implicit. The EU talks constantly about AI sovereignty, but its flagship champion is France's Mistral, and Mistral does not compete at the frontier.
Inviting an American company onto European soil is a different theory of sovereignty entirely one that prizes guaranteed access over domestic ownership. Those two ideas do not fit together easily.
