Anthropic Disables Fable, Mythos After US Directive

Anthropic has disabled its new artificial intelligence models, Fable 5 and Mythos 5. This comes after a US directive to block access for foreigners. The Commerce Department’s action followed a claim from another company that it could "jailbreak" these models. Anthropic, known for its Claude chatbot, stated this directive forced them to shut down the models for all customers, not just foreign ones. The US government communicated this through a notice from Commerce Secretary Howard Lutnick to Anthropic CEO Dario Amodei. Anthropic said the letter didn't specify national security concerns but implied a method to bypass Fable 5's safeguards was known. The company stated it had worked with the US government and other organizations on Fable's safeguards, finding them more effective than previous models, with no "universal jailbreak" found by testers. The US administration reportedly tried to persuade Anthropic to pause the launch before issuing the export control letter. This directive means Anthropic will need a license for the export or transfer of these models, facing penalties if they don't comply. This situation highlights the growing link between AI systems and national security concerns.