Anthropic restricts global access to AI models following US national security directive

Anthropic has suspended access to its Fable 5 and Mythos 5 artificial intelligence models for all foreign nationals, complying with a directive from the United States government that cites national security concerns. The San Francisco-based firm confirmed it received the order on Friday evening at 5:21pm ET, shortly after futures markets had closed. In a statement, the company noted it temporarily blocked access for all customers to ensure strict adherence to the new rules, despite disagreeing with the government’s assessment of the associated security risks.

The directive prohibits foreign nationals, including employees, from accessing the models whether they are located inside or outside the United States. The US government believes it has identified a method of bypassing, or 'jailbreaking', the Fable model. However, Anthropic stated that its own internal tests revealed only a small number of already-known bugs that constituted "minor vulnerabilities." The firm argued that the government’s findings should not necessitate the recall of a commercial model deployed to hundreds of millions of users.

This move marks a significant escalation in the breakdown of relations between Anthropic and the Trump administration. Tensions have grown following the company’s refusal to permit the military use of its technology for domestic surveillance and autonomous weapons systems. In response, the Pentagon has placed Anthropic on a supply chain blacklist, set to take effect later this year, which could severely limit the company’s federal contracts. This export ban represents the second major action by the administration against the AI firm in just a few months.

The scope of the order has drawn scrutiny for its breadth. The New York Times described the directive as "unusually expansive," noting that it could prevent Anthropic’s employees in allied nations, including Canada and the United Kingdom, from using the AI models. Earlier this month, US President Donald Trump signed an executive order requiring federal vetting of the most advanced AI systems for national security risks for up to a month before public release. Until now, US export controls have primarily targeted AI chips and hardware rather than limiting foreign access to the models themselves.

The regulatory intervention raises fresh questions for investors regarding the company’s ability to maintain its technological edge and navigate increasing political friction. The block could potentially hinder Anthropic’s plans for a public listing or initial public offering as early as autumn 2026, with a valuation approaching $1 trillion. The firm warned that if such rules were applied across the sector, it would essentially halt all new model deployments for frontier model providers.