AI company Anthropic has suspended access to its new models Fable 5 and Mythos 5 following an export control directive from the US government citing national security concerns. The decision has reignited debate over who can use the company’s latest AI systems.
Directive includes foreign nationals
According to reports, the federal government requested that access to Fable 5 and Mythos 5 be blocked not only for users outside the US but also for foreign nationals located both inside and outside the country. This scope reportedly includes company employees as well.
The decision signals that commercial AI models may face tighter scrutiny under national security and export control frameworks. Washington is particularly monitoring how advanced models could be used in military, cybersecurity, and critical infrastructure sectors.
Reason cited: Alleged bypass of security measures
US officials have not provided detailed explanations about the national security threat underlying the decision. However, the government believes a method enabling the bypass of security measures in the Fable 5 model has been identified.
Therefore, the decision is significant not only for model access but also for AI companies’ security testing, red teaming efforts, and relationships with government agencies. How well companies can enhance model capabilities while limiting misuse risks has become a key focus for regulators.
Previous tensions with the Pentagon
Anthropic’s refusal to make its Claude model available for "all lawful military purposes" and related security concerns had previously caused disagreements with the Pentagon. The Fable 5 and Mythos 5 decision shows this tension has now resulted in concrete access restrictions on new models.
The company’s compliance with the directive and suspension of access demonstrates that government orders can have direct operational impacts on major AI providers. The process is expected to spark new debates on model security, export controls, and foreign user access going forward.
