Anthropic has suspended its new Fable 5 and Mythos 5 artificial intelligence models following a directive from US national security authorities. The company stated it must abruptly disable the tools for all customers to ensure regulatory compliance just days after their public release.
US officials reportedly identified a method to bypass safety restrictions on the Fable 5 model. While Anthropic acknowledged the demonstration, the company maintains the exposed vulnerabilities are minor and discoverable in other publicly available systems without specialized exploits.
This suspension occurs amidst heightened friction between Anthropic and the Trump administration. Defense Secretary Pete Hegseth recently designated the firm a supply chain risk, a label historically reserved for adversarial foreign entities. Anthropic is currently suing the Pentagon over this classification, though a federal judge has temporarily blocked enforcement of the ban.
Prior to this order, industry leaders had expressed concern regarding the model's advanced capabilities. Anthropic previously described the system as potentially too powerful for unrestricted public access due to its ability to identify cyber vulnerabilities, prompting a limited private testing phase before the general rollout.