U.S. cutoff of Anthropic AI access fuels sovereignty push
European officials and AI experts say the move exposed reliance on U.S. systems and renewed calls for domestic models, cloud capacity and compute.
By Daniel Okafor · Business Editor
4 min read
The U.S. government’s decision to block global access to Anthropic’s most powerful AI models has set off fresh demands for countries to control more of the technology they rely on. Fortune reported that the move marked the first major use of a U.S. power that critics had warned could cut foreign users off from American AI systems.
European politicians cast the shutdown as proof that dependence on U.S. technology has become a strategic risk. Fortune reported that officials revived calls for “sovereign AI,” a term they use for national or regional control over AI models, computing infrastructure and data.
The concern rests on a long-running imbalance. The European Commission and European Parliament say the EU depends on non-EU countries for more than 80% of its technology and 70% of its cloud computing. Epoch AI, an AI research firm, estimates that the U.S. and China control about 90% of global AI computing infrastructure.
Several European governments had already moved to reduce reliance on U.S. providers, according to Fortune. The European Parliament replaced Google with French search engine Qwant over data-collection concerns, while Germany, France and the Netherlands have shifted Microsoft off parts of public-sector infrastructure.
In Britain, Member of Parliament Al Carns said hospitals, companies and researchers had lost access to Fable 5, Fortune reported. Tom Tugendhat, a former U.K. security minister, said the episode showed national security now depends more on software than traditional weapons, and he criticized Britain’s regulatory approach for stressing safety over competitive AI capacity.
Kanishka Narayan, the U.K.’s AI minister, linked the issue to defense. He told Fortune that leading AI models are now used in drones, counter-drone systems and cyber security, making domestic AI capability central to national security.
French and European officials pointed to Mistral as the region’s most credible AI company, though Fortune reported that its models trail leading U.S. systems. Fortune also reported that Mistral is in talks to raise $3.5 billion at a $23.2 billion valuation.
Former French prime minister Édouard Philippe said the shutdown showed AI has become infrastructure on the level of electricity or the internet. Bruno Retailleau, a French 2027 presidential candidate, said the episode should wake up governments to the risk of being cut off from technology supplied by others.
The EU had released a response shortly before the shutdown. On June 3, the European Commission announced a European Technological Sovereignty Package covering cloud computing, AI, semiconductors and open source, including Chips Act 2.0 and the Cloud and AI Development Act.
The Commission said the cloud and AI proposal aims to triple EU data center capacity over five to seven years. Commission President Ursula von der Leyen said Europe cannot rely on others for technologies that support hospitals, energy grids and secure services.
Fortune reported that the package targets about €422 billion, or $490 billion, in investment over a decade, though the Commission has not finalized the funding plan. Critics cited by Fortune said enforcement would depend heavily on member states and that only about 10% of cloud contracts would face a strong European sovereignty standard.
Sandra Wachter, a professor of technology and regulation at the University of Oxford, told Fortune that Europe should consider smaller, more efficient AI models rather than trying to match U.S. frontier systems at any cost. She also said alliances with like-minded countries could offer Europe a more durable path.
The debate is not confined to Europe. Fortune reported that AI sovereignty and a possible coalition of “middle powers” were major topics at February’s AI Impact Summit in New Delhi. Canadian AI pioneer Yoshua Bengio told Fortune that AI sovereignty is tied to democracy and the risk of a world dominated by two technological powers.
Jonathan Iwry, a fellow at the Wharton Accountable AI Lab, told Fortune that the U.S. may have national security reasons for action but risks weakening trust in American technology. He said unilateral control could push allies to reduce their dependence on U.S. AI over time.
This story draws on original reporting from Fortune.