After a negotiation process that dragged on for two weeks, Anthropic’s Mythos 5 can be redeployed to a small, approved set of organizations. The broader, public-facing Fable 5 model still appears stuck, with no rollout timeline set.
By June 26, Anthropic finally had a path back into the cybersecurity fight—though it comes with tight locks on who can use its strongest model.
A government letter dated June 26 and sent by Commerce Secretary Howard Lutnick to Anthropic co-founder Tom Brown—who had been leading negotiations—tells the company it can move Mythos 5 again. The letter describes a “revision to the license requirements. ” grounded in the fact that Anthropic has “worked with the U.S. government to address risks” tied to both Mythos 5 and Fable 5.
Anthropic spokesperson Danielle Ghiglieri said the company received notice that Mythos 5. its “strongest cybersecurity model. ” “can be redeployed to a small group of cyber defenders and infrastructure providers.” Anthropic says it is working to provision “the approved set of providers” and “restore their access to Mythos 5 as quickly as possible.” She added that the company is “pleased to see this progress” and will keep working with the government to expand access to Mythos 5 and to make Fable 5 available for general use again.
But the most important detail for many teams is what’s still not changing. The U.S. government did not lift the export control directive it hit Anthropic with two weeks ago—one that barred any foreign national from accessing either Mythos 5 or Fable 5. including Anthropic’s own employees who aren’t U.S. nationals. Instead. the government carved out an exception for Mythos 5. approving access for a select group of organizations in the same way it did for OpenAI’s GPT-5.6.
Under that exception, both Anthropic employees who aren’t U.S. nationals and members of the approved organizations who aren’t U.S. nationals are cleared to access Mythos 5, according to the letter. Lutnick’s message frames it as progress with safeguards: “These efforts have yielded significant progress.” He said Anthropic committed to work with the government on “protocols and standards and releases for [Mythos-class models].” And he concluded that “appropriate safeguards are in place to permit certain trusted partners to access the Claude Mythos 5 Model.”.
The letter also makes clear this could still shift. Lutnick wrote, “All other requirements of the June 12 letter remain in effect until further notice.” He added, “I reserve the right to reevaluate and adjust the scope of license requirements on [Mythos 5 and Fable 5], should circumstances change.”
For Anthropic, the distinction between models is now the story. Mythos 5 is moving—at least for approved participants. Fable 5 is still in limbo, with no apparent timeline for a rollout agreement.
Pressure had been building on the Trump administration to change the case-by-case regulatory approach it recently adopted. The frustration wasn’t just about access itself. Competitors’ cybersecurity-focused models kept improving, and some were pulling ahead on certain cybersecurity benchmarks. Inside the U.S. AI industry, concerns also grew about what China could achieve while major U.S. labs were sidelined. In addition, top U.S. government departments, including the National Security Agency, had lost access to Mythos 5.
There’s a sharp parallel here: the deal Anthropic has now resembles the structure tied to OpenAI’s GPT-5.6—limited preview access for approved organizations. including trusted enterprises and the U.S. government itself. Both AI labs are hoping for broader availability soon. whether for more enterprise customers or for public access. including Anthropic’s Fable 5. But the timing is ultimately in the hands of the Trump administration.
OpenAI, for its part, argued against making this kind of access process the default. In its GPT-5.6 blog post. OpenAI wrote: “We don’t believe this kind of government access process should become the long-term default. It keeps the best tools from users. developers. enterprises. cyber defenders. and global partners who need them.” The company said it is taking the step “because we believe it is the strongest path to broader availability in the coming weeks. ” while it works with the Administration on a cyber Executive Order framework and a repeatable process for future releases.
So for now, the cybersecurity world gets a partial return. Mythos 5 can be redeployed to a narrow set of trusted organizations. The larger question—when more people can use these tools. and how “revision to the license requirements” becomes a stable system rather than a temporary exception—hangs on what the administration does next.
Anthropic Mythos 5 Fable 5 export control license requirements Howard Lutnick Tom Brown cyber defenders infrastructure providers cybersecurity model GPT-5.6 Claude Mythos 5
So they can use it again but only for like approved people? Cool, love that for everyone.
I don’t get it… they waited 2 weeks just to lock it down more? Feels like the government is just slow-walking access and calling it “negotiations.”
Foreign employees can’t use it even internally?? That’s messed up. But also like, if it’s cybersecurity, wouldn’t other countries need it too? This sounds like politics pretending to be tech.
Fable still waiting?? Typical. They’re all “working with the government” while the rest of us just twiddle our thumbs. Also export controls but then some random small group gets it like it’s normal? Idk, sounds like OpenAI’s deal is just being copied.