Anthropic has complied with a June 12th government order to block all foreign access to its newest AI models, Fable 5 and Mythos 5, just three days after their June 9th launch. The company shut off access entirely for all users—both inside and outside the US—to meet the directive, which included barring access to foreign-national employees of Anthropic itself.
According to the Wall Street Journal, Amazon's security research triggered the order. Amazon researchers reportedly found that through targeted prompts, Fable 5 could be made to generate information potentially useful for cyberattacks. Amazon CEO Andy Jassy conveyed these findings to the White House, leading swiftly to the export control directive.
Anthropnic announced Fable 5 as its most powerful model yet, claiming it "shows exceptional performance in software engineering, knowledge work, and vision," with advantages growing on longer, more complex tasks. Mythos 5 is built on the same underlying model but with "safeguards lifted in some areas," the company said.
In a statement, Anthropic said it was complying with the order but disagreed with the rationale. "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 company said. Anthropic also noted the government "did not provide specific details of its national security concern" and claimed any evidence of vulnerability was delivered verbally, not in writing. The company further stated that similar vulnerabilities are available in other models, including GPT 5.5.
Semafor reported that the White House's decision was also driven in part by fears that a China-linked group may have accessed Mythos 5—a potential national security risk if true. However, the White House has not confirmed this account. In a post on X, Trump adviser David Sacks focused instead on reported jailbreak capabilities, which Anthropic has denied. An Anthropic spokesperson told Semafor that the government did not mention China during its discussions about the export controls.
The shutdown comes as Anthropic was already navigating a separate dispute with the Pentagon. The incident has prompted commentary abroad about the risks of depending on US-controlled AI infrastructure, with some arguing it reinforces the case for non-American companies and governments to develop their own frontier AI capabilities.