US lifts export ban on Anthropic's powerful Fable, Mythos AI models

Anthropic announced on Tuesday that the US Commerce Department has removed export restrictions on its Fable and Mythos AI model, giving the tech giant a major boost in the competitive landscape.

The announcement comes after the Trump administration ordered the US-based artificial intelligence company to suspend the public access to Anthropic’s power Mythos-class AI models over the concerns of national security.

The Trump administration has adopted an assertive oversight strategy for frontier AI models, citing concerns that advanced systems could be exploited by military intelligence agencies in nations such as China and Russia.

In a recent development, Anthropic has reached a deal with the Trump administration to restore access to Fable, which is a public version of powerful Mythos.

"We've received notice that the Department of Commerce has lifted export controls ​on Claude Fable 5 and Mythos 5. We'll begin restoring access tomorrow," Anthropic said in an X post.

A letter was sent by US Commerce Secretary Howard Lutnick to Anthropic, mentioning the withdrawal of restrictions while adding Anthropic had “agreed to proactively detect and address security risks associated ​with the models.”

According to Lutnick, the department "reserves the right to reevaluate the decisions made in this letter and the necessity of reimposing a license requirement, should circumstances change or should Anthropic fail to adhere to its commitments".

Moreover, as a part of deal Anthropic would work with the US government to set the protective protocols for Fable, Mythos and future models. In the case of any breach and malicious activity, the company would inform the administration.

As reported by Wall Street Journal, Anthropic implemented a new safeguard to block workarounds identified by Amazon researchers that previously allowed users to evade safety guardrails to perform cyberattacks. Anthropic claims this new measure stops 99 percent of such evasion attempts.

Additionally, Anthropic is also working with Google, Microsoft and Amazon to develop a “consensus framework” aimed at defining and measuring the grave nature of “jailbreaks”. It will also standardize how developers should respond.

The lifting of restrictions comes after the previous week the US allowed the Anthropic to release Claude Mythos 5 AI to over 100 trusted American organizations.

The security experts are still concerned. Isaac Harris, executive director of the Frontier Security Institute, a ​nonprofit focused on AI ⁠and national security, questioned the US capabilities to handle the technology equipped with unprecedented capabilities coming from China.