The Trump administration has cleared the way for Anthropic Mythos 5 access to resume at more than 100 US companies and government agencies, two weeks after an abrupt export control directive forced the AI lab to pull both Mythos 5 and Fable 5 from the market entirely. Call it a partial thaw: Fable 5 remains off the table for now.
How Anthropic Mythos 5 Access Was Blocked in the First Place
The Commerce Department’s Bureau of Industry and Security issued the original directive on 12 June 2026 at 5:21pm ET, ordering Anthropic to suspend all access to both models by any foreign national, whether inside or outside the United States, with no advance notice and no specific explanation of the national security concern, according to Volkov Law’s analysis of the directive.
The trigger was a reported bypass of the models’ guardrails by security researchers. Mythos 5, the stronger of the two, had been released just days before the ban with what were described as enhanced protections. Those protections, the government concluded, had been circumvented. At least one researcher who saw the underlying work disputes that framing altogether, describing it to TechPolicy.Press as defensive research rather than a jailbreak. The distinction matters, but the ban landed either way.
A group of cybersecurity experts sent a letter urging a reversal on 15 June 2026, per Reuters. The pressure appears to have had some effect.
What the Friday Directive Actually Says
Commerce Secretary Howard Lutnick wrote to Anthropic’s chief compute officer Tom Brown on Friday, stating: ‘I have determined that appropriate safeguards are in place to permit certain trusted partners to access the Claude Mythos 5 Model,’ according to Semafor, which saw the letter.
Lutnick’s letter went further than lifting an access ban. According to TechPolicy.Press, it stated that ‘a license will no longer be required to export, reexport, or in-country transfer’ the Mythos 5 technology to the specific listed entities, their foreign national employees, and Anthropic’s own non-American staff. That last group had been caught in the original ban, too.
The letter cited ‘significant progress’ in talks since 12 June and noted that ‘Anthropic has committed to work with the US government on protocols and standards and releases’ for its models going forward.
The more than 100 organisations cleared for Anthropic Mythos 5 access include many Fortune 500 companies, according to a source who spoke with Reuters. A Commerce Department spokesperson called it a demonstration that ‘America remains the global leader in AI while safeguarding our security.’
Anthropic acknowledged the development publicly on X: ‘Since June 12, we’ve been working closely with the US government to restore access to Claude Mythos 5 and Fable 5. Today, the government notified us that Mythos 5, our strongest cybersecurity model, can be redeployed to a set of US organizations that operate and defend critical infrastructure. 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.’
Fable 5 and the Glasswing Backstory
The Friday directive did not address Fable 5. Conversations between Anthropic and the government were expected to continue into the weekend, focused on restoring that model’s availability, according to a source familiar with the discussions who spoke with CNN.
The two models have a specific relationship worth understanding. According to Anthropic’s official product page, Mythos 5 is the same underlying model as Fable 5 but with safeguards lifted in certain areas, designed specifically for cybersecurity work. Mythos 5 was initially deployed through Project Glasswing, in collaboration with the US government, beginning in April 2026 as an upgrade to Claude Mythos Preview. That first Mythos-class model reached only a limited group of cyber defenders and critical software infrastructure providers before the broader release that preceded the ban.
The timing of Friday’s move was not lost on anyone watching the space. Hours before Lutnick’s letter, OpenAI separately announced it would release its GPT-5.6 family in phases at the federal government’s request, according to NBC News. Two of the biggest AI labs in the country, both navigating government oversight on the same Friday afternoon.
The question now is whether the Mythos 5 restoration represents a workable template or a one-off. The government has signalled it wants Anthropic to commit to ongoing protocols before broader access resumes. Whether that commitment satisfies both sides, and whether Fable 5 follows by next week, will tell us rather more about where AI export controls actually land.
