White House export ban forces Anthropic to pull models

White House – Anthropic says U.S. export controls barred foreign nationals from using its latest AI models, Fable 5 and Mythos 5, forcing the company to take them offline to comply. The dispute comes after tense phone calls involving Anthropic CEO Dario Amodei, Treasury Sec

On Friday, the White House imposed strict export controls on two of Anthropic’s newest AI models, Fable 5 and Mythos 5—then, days later, Anthropic made the decision that the brinkmanship itself had already hinted at.

The company pulled the models offline to “ensure compliance,” Anthropic said, after the U.S. government banned their use by foreign nationals.

The legal directive cited “national security authorities,” Anthropic said, but did not spell out the specific concern. In the weeks and months when AI release schedules and national-security boundaries are getting tighter. the shortness of that explanation—paired with the speed of the takedown—has raised the temperature around what the government believes is at stake.

Anthropic’s pushback is pointed. In a blog post. the company wrote that the Trump administration appears to believe that Fable 5—released days earlier—could be bypassed or jailbroken. Anthropic disputed that premise. saying the “potential jailbreaks that have been disclosed to us are either entirely benign responses or are minor findings that provide no Mythos-specific uplift.”.

The White House and Anthropic did not respond to requests for comment by Business Insider on Monday.

Before Fable 5 and Mythos 5 were pulled. the road to the showdown moved through conversations that sounded less like a formal process and more like an urgent containment effort. In the frantic 24 hours before the crackdown. senior Trump administration officials tried to persuade Anthropic to voluntarily take Fable 5 offline. after concerns raised by Amazon CEO Andy Jassy. Politico reported. citing unnamed administration officials.

That effort played out through tense phone calls between Anthropic CEO Dario Amodei and administration officials including Treasury Secretary Scott Bessent and White House Cyber Director Sean Cairncross. After the order. senior Anthropic officials traveled to Washington. DC. to meet with White House officials in an attempt to resolve the issue. according to reports.

For Anthropic employees and anyone watching the company’s product roadmap, the export controls have an immediate, personal meaning. The company said the government’s export controls barred access to Fable 5 and Mythos 5 for any foreign national—whether inside or outside the United States—“including some of Anthropic’s non-US employees.” Anthropic added that access to all other Anthropic models will not be affected.

The dispute is now sitting at the intersection of AI competition, cybersecurity risk, and U.S. policy power. Industry insiders say this could set an early precedent for how far White House officials will go to block releases they see as a national security risk.

Outside the U.S., the issue also lands with broader consequences: the restrictions have renewed worries about the risks tied to relying on American AI models.

Cybersecurity leaders have also weighed in. with dozens of AI industry and cybersecurity leaders—including specialists from Nvidia and Adobe—signing an open letter urging the government to lift the export controls on Anthropic’s models. The letter. addressed to Commerce Secretary Howard Lutnick and White House Cyber Director Cairncross. said: “Anthropic has built multiple protections into the Fable model to prevent its use for cyber offensive uses. These protections were so aggressive as to be the source of humor in the cyber community on launch day.”.

It continued: “To pull the best capabilities away from defenders without a good reason when our adversaries are rapidly advancing is dangerous.”

Not everyone is convinced the debate can be reduced to a single verdict, either. The letter acknowledged a wider conversation in which “two things are true at once,” even as signatories urged the administration to reverse the restriction.

What remains unclear is the most consequential part of the story. Key questions are still unanswered: exactly what evidence led the government to intervene, the full extent of what the White House order means for foreign nationals working at Anthropic, and whether the effort in Washington will work.

Until those answers arrive, Anthropic’s Fable 5 and Mythos 5 sit offline—an abrupt pause that signals just how quickly AI release decisions can become political decisions, too.

Anthropic White House export controls Fable 5 Mythos 5 Dario Amodei Scott Bessent Sean Cairncross Howard Lutnick AI models national security cybersecurity Nvidia Adobe