AI startup Anthropic has immediately stopped offering its highly capable AI models, Claude Mythos 5 and Fable 5, to all customers. This action follows a directive from the US government, received on December 12th, citing national security concerns. The order prohibits access to these models by foreign entities, both domestically and internationally. Anthropic stated the government instructed them to block access from individuals of all nationalities. These models are known for their advanced capabilities in identifying software vulnerabilities. The company complied with the order, ceasing provision of the AI tools at 5:21 PM EST on December 12th (6:21 AM JST on December 13th). This move highlights increasing US government scrutiny of advanced AI technologies and potential security risks.