Anthropic will disable access to Mythos and Fable models to comply with the Trump Administration's export control

In a significant move, Anthropic has announced it will suspend access to its advanced AI models, Fable 5 and Mythos 5, in response to new export control regulations imposed by the U.S. government. This decision comes after the company received a notification late Friday, citing national security concerns that affect both domestic and international users. The government directive explicitly targets foreign individuals and entities, including Anthropic's own employees who are not U.S. nationals. To comply with these regulations, the company stated that it must disable access for all users to ensure adherence to the order. Notably, the letter from the government did not elaborate on the specific national security issues at stake. Anthropic believes the government's apprehensions stem from a potential vulnerability in Fable 5 that could allow unauthorized access, often referred to as a "jailbreak". However, the company contests the seriousness of these concerns, arguing that the technique in question is limited and revolves around well-known vulnerabilities that can also be detected by other publicly available AI models. This development adds another layer to the ongoing tensions between Anthropic and the Trump administration regarding AI safety and government oversight of advanced technologies. Earlier this year, the Pentagon had flagged Anthropic as a supply chain risk, particularly after the startup pushed for restrictions on the use of its AI models in defense-related applications. Currently, two lawsuits challenging this supply chain designation are still pending. While Anthropic is in compliance with the government's order, it has expressed disagreement with the findings. The company has not specified when access to the Fable 5 and Mythos 5 models will be terminated, but reassured users that other existing models will remain available.