Anthropic will restore access to its most powerful AI models on Wednesday after the US government abruptly lifted an export ban imposed just weeks ago over national security fears. The company said in a statement that the restrictions on Claude Fable 5 and Mythos 5 had been removed following notification from the Department of Commerce.
The two models were suspended on 12 June over concerns they could be used by hackers to exploit weaknesses in computer systems. Commerce Secretary Howard Lutnick wrote in a letter seen by the BBC that Anthropic had addressed the risks. “Anthropic has agreed to proactively detect and address security risks associated with the models,” Lutnick wrote, adding that the firm had also agreed to collaborate on future releases and alert the government of any malicious activity. The department reserves the right to reconsider the decision if necessary, he said.
“US lifts export ban on Anthropic's AI tools after company agrees to address security risks.”
Fable 5 and Mythos 5 are built on Anthropic's Claude platform, a rival to OpenAI's ChatGPT and Google's Gemini. Fable 5 is designed for consumers, capable of deep reasoning and complex tasks, while Mythos 5 targets businesses and cybersecurity experts and can identify and exploit vulnerabilities in computer code. Both were released on 9 June.
At the time of the suspension, Anthropic said US authorities had not pinpointed specific concerns about its technology. “Our understanding is that the government believes it has become aware of a method of bypassing, or ‘jailbreaking’ Fable 5,” the company said then, referring to slipping past safety restrictions. “However, 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.”
The Commerce Department said in its letter that Anthropic has now agreed to detect and address security risks proactively. The BBC has contacted the department for further comment.