Anthropic announced on Friday that it has received US government approval to restore access to its Mythos 5 AI model for a limited set of American organizations, partially reversing a sweeping ban imposed earlier this month. The decision comes after weeks of negotiations between the company and Washington, but leaves European users and allies questioning the continent's dependence on US-controlled AI technologies.
The Trump administration's 12 June directive prohibited foreign nationals—including Anthropic employees—from using the firm's Fable 5 and Mythos 5 models, citing national security risks. In response, Anthropic had to cut off all customer access to comply. Now, the company says it can redeploy Mythos 5, its strongest cybersecurity model, to US organizations that operate and defend critical infrastructure.
"We’re restoring access for these organizations quickly, and we’re continuing to work with the government to expand access to Mythos 5 and make Fable 5 available for general use again," Anthropic stated on Friday.
European Reactions and the Push for Tech Sovereignty
The initial ban sent shockwaves through European capitals, where officials saw it as a stark reminder of the risks of relying on US-based AI infrastructure. Benjamin Haddad, the French minister delegate for Europe, responded at the time by urging Europe to "invest more, support its innovators, and equip itself with the means to master the technologies that will determine power in the 21st century."
His comments echoed broader concerns within the EU about strategic autonomy in critical technologies. The European Commission has since accelerated discussions on a joint AI framework, with several member states pushing for a coordinated response to US export controls. The incident has also revived debates about the EU's earlier warnings that such restrictions must not discriminate against European entities.
For now, Anthropic's partial restoration of Mythos 5 does little to address European grievances. The model remains inaccessible to non-US organizations, including those in EU member states that rely on advanced AI for cybersecurity, research, and industrial applications. The company has not provided a timeline for broader international availability.
The situation underscores a growing transatlantic rift over AI governance. While the US prioritizes national security and domestic control, European policymakers are increasingly calling for a more balanced approach that ensures fair access and fosters homegrown innovation. The EU's own AI Act, still under final negotiation, aims to create a regulatory environment that could support European alternatives to US-dominated platforms.
Meanwhile, the geopolitical stakes are high. As Ukraine and Russia exchange strikes and EU accession talks advance, the need for secure, sovereign AI capabilities becomes more pressing. European leaders are watching closely to see whether Washington will extend similar access to allied nations or continue to tighten controls.
Anthropic's statement suggests ongoing dialogue with the US government, but the path to full reinstatement of Fable 5 and Mythos 5 for general use remains unclear. For Europe, the episode serves as a catalyst for a long-overdue conversation about technological independence in an era of great-power competition.
