US Government Directs Anthropic to Suspend Advanced AI Models
The U.S. government has issued an emergency export control directive, compelling AI company Anthropic to immediately suspend access to its flagship Fable 5 and Mythos 5 artificial intelligence models. This action, taken on June 12, cites national security concerns, specifically related to a "jailbreak" exploit that could allow the models to scan for software vulnerabilities. To guarantee absolute compliance with the order, Anthropic disabled both models globally, affecting all customers, including domestic U.S. citizens and foreign nationals.
The United States government has issued an emergency export control directive, instructing Anthropic to immediately suspend access to its Fable 5 and Mythos 5 artificial intelligence models. This directive, which arrived abruptly on June 12, cites national security authorities as the basis for the action.
Anthropic was directed to suspend access for all foreign nationals, whether they are located inside or outside the United States, including foreign national employees of Anthropic. To ensure complete compliance with the government's order, Anthropic took the measure of completely disabling both models for all customers worldwide, including those in the U.S.
The restriction reportedly stems from a codebase "jailbreak" exploit. This vulnerability allows users to potentially utilize these AI models to scan for software vulnerabilities. The US government cited national security concerns, though little initial documentation was provided alongside the directive.
Anthropic has reportedly pushed back against the directive, arguing that the identified flaw is minor and that similar vulnerabilities are present in competitor models, such as GPT-5.5. The company contended that the flaw should not warrant a commercial blackout of its advanced AI models.
According to Android Authority, this move represents an unprecedented action by the US government regarding advanced AI models.
Advertisement
AdSense slot • inline
