US Government Clears Anthropic to Deploy Mythos 5 AI Model to Select Organizations

Government Lifts Export Block on Powerful AI Cybersecurity Tool

The US government authorized Anthropic to restore access to its Mythos 5 AI model for a select group of organizations, marking a significant de-escalation in tensions between federal authorities and one of the world’s most valuable private companies. Commerce Secretary Howard Lutnick sent a letter to Anthropic on Friday afternoon confirming the decision, according to multiple reports viewed by NBC News and Semafor.

The decision comes two weeks after Lutnick invoked export control authorities, forcing Anthropic to shut down access to both Mythos 5 and its consumer-facing cousin, Fable 5, citing threats to national security. In his Friday letter to Anthropic’s chief compute officer Tom Brown, Lutnick wrote that the government was confident in the guardrails Anthropic had put in place to allow trusted users to access the powerful AI system.

Mythos 5 access will be restored to around 100 organizations that include government agencies and private companies, according to people familiar with the matter. The individuals said the restoration will help ensure the organizations can use the model for defensive cyber purposes. The approximately 100 organizations receiving restored access include government agencies and private companies focused on protecting critical infrastructure.

“I have determined that appropriate safeguards are in place to permit certain trusted partners to access the Claude Mythos 5 Model,” Lutnick wrote, citing “significant progress” in the intense, daily talks between the government and the company since the block went into effect.

New Framework for AI Model Oversight Takes Shape

Lutnick’s letter marks the beginnings of a new regulatory regime that gives the US government control over the release of frontier AI models. Under the new arrangement, a license will no longer be required to export, reexport, or in-country transfer the Claude Mythos 5 Model to entities identified in an annex to the letter and their foreign national employees, or to Anthropic’s foreign national employees.

The Commerce Department did not immediately respond to a request for comment Friday night. However, a department spokesman cited the speed with which the government acted to address its concerns about Anthropic. The framework for overseeing AI is being built on the fly, with many users of the powerful tools watching closely as federal authorities establish protocols.

Anthropic confirmed the development in a post on social media platform X on Friday night, stating that the government notified the company that Mythos 5, its strongest cybersecurity model, can be redeployed to US organizations that operate and defend critical infrastructure. The company added that it is working to provision the approved set of providers and restore their access to Mythos 5 as quickly as possible.

A source close to the company said that Anthropic will continue to hold discussions with the government over the weekend as it seeks to restore access to Fable 5, its most recent and most powerful mass-consumer model. The letter remains silent on Fable 5, though people close to the talks said they are moving toward releasing Fable as well, though that timeline is unclear.

OpenAI Faces Similar Restrictions on Latest Model Release

Lutnick’s Friday move that allows Anthropic to restore limited access to Mythos 5 comes hours after Anthropic competitor OpenAI announced it would release its latest and most advanced series of models, the GPT-5.6 family, in phases per the federal government’s request. OpenAI CEO Sam Altman wrote on X that the staggered debut was “bad news,” as the company had instead planned on a wider, open-access launch.

In a blog post announcing the new models, OpenAI wrote that the company had shared the list of trusted partners with the government before the models’ launch. The AI giant said it would work with the administration to develop a more robust framework to vet and clear future models for public release. The move underscores how federal authorities are increasingly asserting control over the deployment of advanced AI systems.

The recent action against Anthropic comes after experts fretted for months that Mythos could pose a security risk. They warned the model could give hackers and other bad actors access to AI capable of finding and exploiting vulnerabilities exceptionally quickly. The federal government’s concerns centered on the potential for malicious actors to jailbreak the models for harmful purposes.

Critical Infrastructure Protection Drives Restoration Decision

Earlier this month, Anthropic disabled customer access to both Mythos and Fable to comply with the US government’s order to suspend all use by foreign nationals, including Anthropic employees themselves. The export ban and subsequent negotiations with Anthropic underscored the lack of a consistent regulatory framework around AI, even as the technology advances rapidly and the US tries to stay ahead of global competitors like China.

Until recently, the government has taken a largely hands-off approach to regulating artificial intelligence, seeking to keep US companies ahead of Chinese competitors in particular. The sudden invocation of export controls marked a dramatic shift in policy, catching both industry observers and company executives by surprise. The two-week negotiation period that followed involved intense, daily talks between government officials and Anthropic representatives.

“Anthropic has committed to work with the U.S. government on protocols and standards and releases” for its models, Lutnick wrote in his letter to the company.

Conversations between Anthropic and the government are expected to continue into the weekend, with an eye to restoring access to Fable, as well, a source familiar with the discussions confirmed. The ongoing negotiations highlight the delicate balance federal authorities are attempting to strike between maintaining US leadership in AI development and addressing legitimate national security concerns.

Industry Watches as Regulatory Precedent Forms

The Mythos 5 case represents the first major test of how the US government will regulate the most advanced AI systems. Though the leaders of frontier labs have bridled at the possibility of losing time in the intense global AI race, federal officials have emphasized the need for careful oversight of systems with potentially dangerous capabilities. The speed of the resolution suggests both sides recognize the importance of finding workable solutions.

Industry analysts note that the selective release approach could become a template for future AI deployments. By restricting access to vetted organizations focused on defensive applications, the government aims to capture the security benefits of advanced AI while minimizing risks. The arrangement allows critical infrastructure providers and government agencies to leverage Mythos 5’s cybersecurity capabilities without enabling broader access that could facilitate malicious uses.

As Anthropic works to restore Mythos 5 access to approved organizations and continues negotiating over Fable 5, the AI industry confronts a new reality. Federal oversight of frontier AI models appears poised to become a permanent feature of the technology landscape, requiring companies to work closely with government authorities to deploy their most capable systems. The outcome of these early cases will likely shape AI regulation for years to come.