Anthropic has disabled its most advanced artificial intelligence models after receiving a United States government order requiring the company to suspend access for foreign nationals, citing national security concerns.
The company said it received the export control directive on Friday and was instructed to restrict access to its newly launched Fable 5 model and the more advanced Mythos 5 system.
According to Anthropic, the order provided limited details regarding the specific security concerns behind the decision.
To ensure compliance, the company said it was forced to disable access to the models for all customers at short notice, while access to its other AI systems remains unaffected.
Anthropic said it understands that government officials believe there may be a method of bypassing, or “jailbreaking,” safeguards designed to prevent Fable 5 from being used to identify software vulnerabilities.
The company said it reviewed information that it believes may have prompted the directive and concluded it involved a limited capability to analyze specific program code and correct errors.
Anthropic argued that similar capabilities exist in competing models offered by other AI developers.
The company maintained that extensive safety testing had been conducted before the release of Fable 5 and said it disagreed with the decision to restrict access based on what it described as a narrow and non-universal concern.
A U.S. official confirmed that the Commerce Department issued the directive suspending access to Fable 5 and Mythos 5 for foreign nationals.
The move represents a significant expansion of U.S. efforts to limit access to advanced AI technologies.
While previous export controls largely focused on semiconductor chips and technical infrastructure, the latest action directly targets access to AI models themselves.
Anthropic said the order affects foreign nationals both outside and inside the United States, potentially including employees born abroad.
The company stated that Mythos-class systems possess advanced cybersecurity capabilities that could be misused if placed in the wrong hands, while also helping identify and fix software vulnerabilities.
Anthropic said it believes the government action stems from a misunderstanding and that it is working to restore access as quickly as possible.
The company warned that applying similar standards across the AI industry could significantly slow future deployments of advanced AI models.