Anthropic’s leadership left Monday’s high-stakes meeting in Washington with the same problem they arrived with: fundamental disagreement with White House officials over whether Claude Fable 5 poses a genuine national security risk. The talks, held between Anthropic executives and the US Department of Commerce under Secretary Howard Lutnick, wrapped without producing a clear path forward, leaving both sides publicly entrenched.
Anthropic’s technical team had already engaged in virtual discussions with White House officials since the Trump administration first reached out on June 12. The company maintains that the government had already tested Fable 5 and approved it before the global release, and that officials have since identified only a narrow jailbreak technique capable of bypassing safeguards built into the model.
Anthropic said it believes the government’s concern centers on a “potential narrow, non-universal jailbreak,” where a user could bypass a cybersecurity guardrail and ask Fable 5 to read a specific codebase and fix software flaws.
“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, adding that applying this standard industry-wide “would essentially halt all new model deployments for all frontier model providers.”
Anthropic has characterized the entire dispute as a misunderstanding and says it is working to restore access as soon as possible.
The political pressure intensified over the weekend. Defense Secretary Pete Hegseth, who earlier this year labeled Anthropic a supply chain risk and barred Claude from the Pentagon, renewed his criticism of the company following the shutdown.
David Sacks, Trump’s special adviser on AI and cryptocurrency through this past March, wrote on X that Anthropic’s response to the directive is “very much at odds with their branding and ethos as a safe AI research” firm.
“The Admin’s hope now is that Anthropic remediates the safety issue, the export control is lifted, and Fable goes back into general release,” Sacks wrote. “It is frankly bewildered that Anthropic hasn’t wanted to comply with safety requests that it previously said were its highest priority.”
The standoff is increasingly viewed as a test case for how aggressively Washington will regulate frontier AI models going forward, with no indication yet of when, or whether, Fable 5 and Mythos 5 will return to public access.
