Anthropic’s statement on the US government directive to suspend access to Fable 5 and Mythos 5

Anthropic reports that the directive appears to be based on a potential “jailbreak” method that could bypass the models’ safety safeguards. A jailbreak is a technique used to manipulate an AI into ignoring its programming or ethical constraints to perform prohibited tasks. While the government expressed concern over cybersecurity risks, Anthropic argues that the identified vulnerabilities are minor and already present in other publicly available AI models. Anthropic's statement on the US government directive to suspend access to Fable 5 and Mythos 5 (image from anthropic.com) The company maintains that its defense-in-depth strategy and rigorous red-teaming make these models safer than many existing industry alternatives. Anthropic expressed disagreement with the government’s decision, stating that the directive lacks transparency and clear technical grounding. Despite this opposition, the models remain offline for all users while the company works with officials to resolve the misunderstanding and restore service. Statement on the US government directive to suspend access to Fable 5 and Mythos 5 Anthropic | anthropic.com Anthropic disables Fable 5 and Mythos 5 after US export directive | letsdatascience.com Anthropic suspends new AI models after government directive | nbcnews.com