It’s to ensure compliance with a government directive citing national security concerns.
Anthroic has disabled all of its customers’ access to Fable 5 and Mythos 5 in order to ensure compliance with an order it received from the government on Friday, June 12. All its other models and its Claude chatbot are not affected. The company said in its announcement that the US government wanted it to suspend all foreign nationals’ access to its newly launched AI models, whether they’re inside or outside the US and even if they’re Anthropic employees, citing national security concerns.
While the US government didn’t specify those concerns, Anthropic believes that it’s because the government heard about a method of jailbreaking Fable 5. The company has just launched the Fable AI model, which was designed to bring many of Mythos’ capabilities to the public, on June 9. If you’ll recall, Mythos is its state-of-the-art cybersecurity model that’s only available to its Project Glasswing partners. Fable’s capabilities “exceed” any previous model Anthropic has launched. It beat Pokémon FireRed during the company’s tests, for instance, while Claude failed to beat Pokémon Red, the original game it was based on.
Anthropic listed the measures it took to ensure that Fable was secure in its post. It said it instituted strong safeguards to “reduce the likelihood that Fable is misused for tasks related to cybersecurity” and added that its “safeguards are so strong that many users have complained that they are overly broad.” The company also explained that any provider cannot possibly ensure perfect resistance to jailbreak attempts, and every model is vulnerable to jailbreaks made especially for it. “We aimed to make jailbreaks either narrow (in the case of non-universal jailbreaks) or very expensive to produce (in the case of universal jailbreaks), and to combine this with thorough monitoring to quickly detect and shut down any successful attacks,” it said about its defense strategy.
The government apparently gave the company verbal evidence for one potential narrow, non-universal jailbreak that an unnamed entity shared with officials. Anthropic promised to share more details over the next 24 hours, but it clarified that it disagrees that a potential jailbreak should be cause for recalling a commercial model.
“As we have stated publicly, we believe the government should have the ability to block unsafe deployments, as part of a statutory process that is transparent, fair, clear, and grounded in technical facts,” Anthropic, which has been vocal about its warnings about the need for more AI oversight, wrote. “This action does not adhere to those principles.”
