Anthropic announced on Friday that it is disabling its two most advanced AI models – Fable 5 and Mythos 5 for all users worldwide. The trigger was a US government export control directive ordering the company to restrict access for foreign nationals, citing national security concerns. Anthropic said it was given no specific details, only verbal evidence of a “potential narrow, non-universal jailbreak” that the government believes could allow Fable 5 to identify software vulnerabilities. The company said it could not selectively block foreign nationals without pulling the models entirely – so a product used by hundreds of millions of people globally has been switched off to solve a problem the government could not precisely define. Anthropic has said it disagrees that a narrow, unverified jailbreak justifies recalling a commercially deployed model, and that if this standard were applied across the industry, it would essentially halt all new frontier model deployments. The order lands against an already fractured backdrop. Anthropic’s relationship with the US government broke down earlier this year after it refused to allow military use of its models for domestic surveillance and autonomous weapons. Washington responded by placing the company on a supply chain blacklist. Anthropic also filed confidentially for a US IPO last month. A Pentagon official posted in support of the directive: “Some things are simply more important than revenue cycles, clickbait, and pre-IPO valuation.”
Source: Reuters
