The myth of the dangerous abilities of the Mythos 5 model.

The United States has recently imposed restrictions on access to advanced artificial intelligence models developed by Anthropic, a prominent tech company known for its cutting-edge language models. These measures have triggered a ripple effect across international policy discussions, particularly within the European Union, where Austria has proposed bringing Anthropic's operations under EU jurisdiction. This move comes as part of broader efforts to ensure technological sovereignty and reduce dependency on American technology giants.

In response to the U.S. restrictions, Austria has called on the European Union to consider relocating Anthropic within the European community. The Austrian government, through a letter addressed to the European Commissioner for Technology, Henna Virkkunen, emphasized the importance of ensuring that Europe remains connected to critical innovations. In this communication, Secretary of State for Digitalization, Alexander Proell, highlighted the need for the EU to explore strategic placement and engagement of Anthropic within the union. He argued that such actions would align with the company’s values, market access, capital, and other factors that make it suitable for European integration. However, he did not specify how this relocation might be implemented, acknowledging potential doubts about its feasibility.

Proell further stressed that the central issue is not the difficulty of achieving such a relocation but rather whether Europeans are ready to take charge of their technological future. He questioned whether the EU aims to be architects of its own technological destiny or merely administrators of decisions made elsewhere. This sentiment reflects a growing desire among European nations to assert control over their digital infrastructure and reduce reliance on foreign entities, especially in light of recent regulatory proposals aimed at boosting the development of the European cloud computing, AI, and semiconductor markets.

These developments occur against a backdrop of increasing tensions between Anthropic and the U.S. Department of Defense. Earlier this month, the U.S. government mandated that Anthropic block access to its latest models—Fable 5 and Mythos 5—for non-U.S. citizens. According to official statements, these restrictions were linked to concerns over the models' capabilities and potential vulnerabilities that could be exploited for malicious purposes. Anthropic complied by restricting access to all clients and working on additional security measures.

However, critics argue that the perceived threat posed by these models is overstated. Similar capabilities exist in both competing large-scale models and smaller models that can be run locally. Additionally, there are ongoing disputes between Anthropic and the U.S. Department of Defense regarding clauses in contracts that prohibit the use of Anthropic's models for surveillance of U.S. citizens and fully autonomous weapons systems, such as drones. The Department of Defense had previously requested the removal of these clauses, which Anthropic refused, citing ethical concerns related to democratic values and the reliability of current models for such applications.

The situation escalated when former President Donald Trump, via social media, ordered all federal agencies to cease using Anthropic's services. Following this directive, Secretary of Defense Pete Hegseth formally labeled Anthropic as a "supply chain risk," effectively barring any business relationships with firms having contracts with the U.S. military. This designation was seen as more of a symbolic gesture than a practical measure, given the lack of concrete justification for such a move. It also highlighted the absurdity of the situation, where the Department of Defense considers unrestricted access to Anthropic's models so crucial that they must be restricted, yet simultaneously views the company as a national security threat requiring such stringent controls.

As the debate continues, Anthropic has not responded to requests for comment regarding Austria's proposal to relocate its operations to Europe. The company's silence underscores the complexity of navigating geopolitical tensions while maintaining operational integrity and innovation. Meanwhile, the European Union faces the challenge of balancing its aspirations for technological independence with the realities of global competition and collaboration in the field of artificial intelligence.