US Clears Anthropic’s Mythos 5 AI Model for Trusted Critical Infrastructure Organizations

The US government has authorized Anthropic to restore limited access to its Mythos 5 artificial intelligence model, allowing selected organizations that defend critical infrastructure to use the cybersecurity-focused system once again.

The decision partially reverses restrictions introduced earlier this month, when the Trump administration suspended access to Anthropic’s Mythos 5 and Fable 5 models over concerns that foreign governments or hostile actors could exploit the technology. While Mythos 5 will return for a carefully vetted group of US organizations, Fable 5 remains unavailable.

Anthropic Begins Restoring Access

Anthropic announced on Friday that federal officials had approved the redeployment of Mythos 5 to trusted organizations responsible for protecting critical infrastructure.

According to the company, it has already started restoring access and plans to expand availability as discussions with the government continue.

“We’re restoring access for these organizations quickly, and we’re continuing to work with the government to expand access to Mythos 5 and make Fable 5 available for general use again,” Anthropic said.

A source familiar with the matter told Reuters that the approval covers more than 100 companies and institutions, including numerous Fortune 500 businesses.

Earlier Restrictions Targeted Foreign Access

The approval follows sweeping restrictions introduced two weeks ago that prevented Anthropic from providing Mythos 5 and Fable 5 to any foreign nationals, regardless of whether they were located inside or outside the United States.

The policy also affected Anthropic employees, forcing the company to disable access to both models almost immediately.

US officials introduced the restrictions after raising concerns that adversaries in countries including China and Russia could misuse the advanced AI systems.

Government Says Safeguards Reduced the Risks

Commerce Secretary Howard Lutnick said Anthropic had worked with federal officials to strengthen protections around the company’s advanced AI models.

In a letter seen by Bloomberg, Lutnick wrote that Anthropic had made significant progress addressing risks associated with the restricted systems.

He concluded that appropriate safeguards now allow selected trusted partners to use the Mythos 5 model.

However, the government has not extended the same approval to Fable 5, which remains unavailable while discussions continue.

Anthropic Continues Negotiations

Anthropic previously believed government officials became concerned after learning that researchers had discovered a way to jailbreak Fable 5 and use it to identify software vulnerabilities.

The company responded by introducing additional safeguards that it said significantly reduced the risk of misuse.

Despite those changes, regulators have only approved the return of Mythos 5.

Anthropic says it will continue working with the US government to broaden access to Mythos 5 while seeking approval to release Fable 5 more widely.

AI Companies Face Growing Government Oversight

The latest decision marks another chapter in an increasingly complicated relationship between Anthropic and the US government.

Earlier this year, the Trump administration classified the company as a “supply chain risk” after Anthropic refused to remove safeguards that prevent its AI models from supporting autonomous weapons or domestic mass surveillance.

Anthropic is not the only AI developer facing tighter controls.

OpenAI also announced on Friday that it would limit access to its new GPT-5.6 models to trusted partners following a request from the US government, highlighting a broader push to restrict access to advanced AI systems while national security concerns grow.