It’s being reported Anthropic has flown staff to Washington DC in a desperate bid to have an export ban removed from the AI company’s most advanced models.
A company source told Axios the company has held virtual meetings with the White House since the administration reached out on Friday with concerns about the potential to “jailbreak” the new Mythos and Fable models.
The company received an export-control directive telling it to suspend access to Fable 5 and Mythos 5 for all foreign nationals.
The precise nature of the national-security concerns were not stated in the order.
According to Reuters, “It is Anthropic’s understanding that the government believes there is a method of bypassing, or ‘jailbreaking,’ a safeguard that would prevent Fable 5 from being used in identifying software vulnerabilities, the company said.”
Relations between Anthropic and the US government have been strained since the company refused to allow the US military to use its AI models for domestic surveillance and autonomous weapons systems.
The government responded by placing the company on a supply-chain blacklist, a directive that is set to take effect later this year.
Reuters notes, “The action also marks a major escalation of U.S. efforts to halt foreign adversaries’ AI capabilities. For years, U.S. export controls have focused on the chips and tools that power AI rather than on restricting foreign access to AI itself.”
On Wednesday, Anthropic called for greater US oversight of AI, including the ability to block models deemed to be a security risk. However, the company has said the directive received on Friday is not reasonable.
“We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people,” Anthropic said in a statement.
“If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers.”
Claude Fable 5 was released this week, with a new tier of abilities Anthropic has described as Mythos-class.”
Although the model has guardrails to prevent its use for hacking, experts have said it could dramatically accelerate sophisticated cyberattacks, especially in sectors like banking.
One Response
Beware of an zionists getting control over AI systems in the government, it would guarantee corruption and great evil.