Anthropic is currently engaged in a high-profile dispute with the U.S. government regarding its latest AI models, Fable 5 and Mythos 5. The Trump administration issued an export control directive on a Friday, ordering Anthropic to suspend access to these models by any foreign national, citing unspecified national security threats. This directive effectively blocked access to the models for all customers worldwide, as Anthropic stated it could not reliably differentiate between foreign and domestic users.
Senior Anthropic staffers held meetings with Trump administration officials in Washington, D.C., on Monday to try and resolve the situation. However, these initial face-to-face talks concluded without a deal, and the restrictions on Fable 5 and Mythos 5 remain in place. Anthropic has characterized the dispute as a "misunderstanding" and is working to restore access to the models as soon as possible, maintaining that a "narrow potential jailbreak" should not warrant recalling a commercial model deployed to hundreds of millions of people.
This incident marks a continuation of strained relations between Anthropic and the government, following an earlier clash with the Department of Defense. The DOD had previously labeled Anthropic a supply chain risk in March, banning defense contractors from using its technology due to perceived threats to U.S. national security. The export control directive regarding Fable 5 and Mythos 5 came just days after their public unveiling, despite Anthropic claiming to have worked with government agencies and received approval to deploy the models beforehand.
Administration officials indicated that the export controls were a "last resort" after hours of attempting to persuade Anthropic to voluntarily pull the models due to security vulnerabilities identified by other tech companies, including Amazon. Anthropic reportedly did not agree to pull the models during these discussions, leading to the Commerce Department's directive. The broader implications of this government action are significant, setting a precedent for future interactions between AI developers and regulators concerning national security and model deployment.
