Anthropic is reportedly working with the US on a policy framework to deal with similar situations in the future.

The US government has partially reversed a ban on Anthropic’s Mythos 5, allowing the company to release the powerful cybersecurity AI model to more than 100 select institutions, including government agencies.

In a letter to Anthropic on Friday (26 June), US Secretary of Commerce Howard Lutnick determined that the company had “appropriate safeguards” in place to repermit Claude Mythos access to certain trusted partners.

“Anthropic has worked with the US government to address risks associated with the covered models,” Lutnick said. “These efforts have yielded significant progress.”

The US government had previously abruptly suspended Anthropic from giving access to Mythos and Fable to “any foreign national, whether inside or outside the US, including foreign national Anthropic employees”.

In its directive to the company two weeks ago, the government cited national security concerns but did not provide specific details, Anthropic said.

The company understood this to refer to a possible jailbreak for Fable 5, but argued that any vulnerabilities “appear[ed] relatively simple” and that they could be discovered by competing, publicly available models. Last Friday’s decision does not ease restrictions on Fable.

“We received notice from the US government that Mythos 5, our strongest cybersecurity model, can be redeployed to a small group of cyber defenders and infrastructure providers,” Anthropic said in a statement. “We are working to provision the approved set of providers and restore their access to Mythos 5 as quickly as possible.

“We are pleased to see this progress and continue to work with the government to expand access to Mythos 5 and make Fable 5 available for general use again.”

Under the new arrangement, licences will no longer be required to export Mythos 5 to identified institutions and their foreign national employees, or Anthropic’s own foreign national employees.

It’s unclear how Anthropic satisfied the government’s national security concerns stemming from a potential Fable jailbreak. A source told Bloomberg that extended conversations over the weekend were aimed at bringing back access to the model soon.

Meanwhile, Anthropic is also reportedly working with the US government to come up with a policy framework to handle similar situations in the future.

Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.

Dario Amodei at the World Economic Forum, 2026. Image: World Economic Forum via Flickr (CC BY-NC-SA 4.0)