The US government plans to lift export controls on Anthropic PBC once the company addresses a "safety issue" that prompted the export suspension of its Fable 5 and Mythos 5 models. 

"The Admin's hope now is that Anthropic remediates the safety issue, the export control is lifted, and Fable goes back into general release," David Sacks, co-chair of the President’s Council of Advisors on Science and Technology, wrote in a post on X. "The Admin wants all of this to happen as soon as possible."

The San Francisco-based AI company had previously warned about the hacking capabilities of its Mythos model and held it back from wide release. Earlier in the week, Anthropic rolled out a public version, called Fable, with what it described as cybersecurity safeguards.

The US government told Anthropic to suspend access to the Fable 5 and Mythos 5 models by any foreign national "whether inside or outside the United States," citing national security concerns, the company said in a statement on Friday. Officials also said they found a way to bypass, or ‘jailbreak,' a safeguard meant to stop the model from being used to identify cybersecurity vulnerabilities.

"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 its post. "If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers."

The move marks one of Washington's most forceful interventions in commercial AI to date. The dispute has quickly become a test of how aggressively the US will enforce safety standards on frontier models 

Anthropic ‘At Odds' With Safe AI

Anthropic has shut off access to both systems for all customers to ensure compliance. The administration, however, has argued that Anthropic demonstrated a "surprising" unwillingness to cooperate with its safety requests. 

"It's been very surprising that Anthropic hasn't wanted to cooperate with a reasonable safety request (i.e. fixing the jailbreak issue)," Sacks wrote. "Anthropic's reaction is very much at odds with their branding and ethos as a safe AI research community."

Researchers at Amazon.com Inc. (AMZN.O) conducted jailbreak research that revealed vulnerabilities in Anthropic's model, according to the Wall Street Journal. The US government and Amazon were in contact about the vulnerability before the controls were imposed, according to people familiar with the matter.

Amazon CEO Raises Concerns

Amazon CEO Andy Jassy was among tech leaders who raised concerns to senior Trump administration officials this week about security risks in Anthropic's most advanced AI models, a person familiar with the matter told Reuters.

"A highly credible trusted partner of both Anthropic and the USG who was testing Fable came forward with a jailbreak of those guardrails," Sacks said. He also criticized Anthropic for saying the jailbreak wasn't serious.

"That is not what the trusted partner and the USG believe," he wrote. "Nor is that kind of minimizing language consistent with Anthropic's brand as the AI safety company."

Some experts who favor export controls on advanced AI models found the Trump administration’s action puzzling, Reuters reported. The decision affects US allies as well as adversaries.

Jimmy Goodrich, a senior fellow at the University of California's Institute for Global Conflict and Cooperation, told Reuters the administration's decision wasn't "well thought-out." The ban prevents even Canadian and British employees at Anthropic from doing research and development, he said.