Austria has urged the European Union to explore establishing Anthropic within the bloc after the United States restricted foreign access to its most advanced AI models. The proposal, detailed in a letter from State Secretary for Digitalization Alexander Proell to EU Technology Commissioner Henna Virkkunen, calls for strategic EU participation in the AI company to safeguard Europe’s access to frontier technology. Proell argued that Europe should not risk losing access to major AI advances because of decisions made outside the region, offering legal certainty, market access, investment, and a values-based environment as incentives. He framed the issue as a test of European technological sovereignty, asking whether the bloc wants to shape its own future or remain dependent on external policy changes.
The move follows a U.S. Commerce Department export control directive that forced Anthropic to suspend access to its Fable 5 and Mythos 5 models globally, including for foreign nationals working inside the United States. Officials cited national security concerns, warning the models could be exploited via jailbreaks to identify or repair software vulnerabilities. The sudden freeze disrupted European companies, researchers, and public agencies that had already integrated the tools, leaving no transition period or immediate workaround.
This development has sharpened the EU’s broader push to strengthen domestic cloud, AI, and semiconductor industries and reduce reliance on U.S. tech firms. Earlier this month, the European Commission proposed legislation to bolster local capacity, and the Anthropic case adds political urgency. The bloc may now pair regulatory oversight with incentives for companies that build or operate inside Europe, potentially altering the direction of its AI Act.
Anthropic also faces business pressure abroad and legal scrutiny at home. JPMorgan removed its Claude models from approved tools for Hong Kong employees due to licensing restrictions, following a similar earlier move by Goldman Sachs for Greater China. In the U.S., a proposed class action in the Northern District of California alleges that the company’s $100 per month Max 5x and $200 per month Max 20x subscriptions delivered significantly less usage than advertised, making it hard for subscribers to predict caps. These challenges unfold as Anthropic recently published a policy proposal calling for government testing requirements, cybersecurity standards, and enforcement for frontier AI systems.
