AI firm Anthropic has limited access to its most powerful AI models following a request from the US government. The company stated the government indicated it had learned of a method to circumvent security measures on Fable 5, a highly capable AI model previously withheld from public release. Fable 5 is based on Anthropic’s Mythos 5 architecture and possesses unprecedented abilities, prompting initial restrictions. While the specific nature of the national security threat remains undisclosed, Anthropic complied with the government’s request to prevent potential misuse. The move highlights growing concerns surrounding the potential risks associated with advanced artificial intelligence. This action underscores increased scrutiny of AI development and deployment by US authorities. Anthropic continues to cooperate with the government on the matter.