Mythos 5 Returns from Anthropic
After a lengthy two-week negotiation with the Trump administration, Anthropic’s Mythos 5 is back in action—at least for a select group of organizations, as revealed in a letter from the government to Anthropic. However, Fable 5, the public-facing model, remains in limbo with no clear timeline for its rollout.
The letter, dated June 26 and sent by Commerce Secretary Howard Lutnick to Anthropic co-founder Tom Brown, indicates a “revision to the license requirements” following Anthropic’s collaboration with the U.S. government to address risks related to both Mythos 5 and Fable 5.
Anthropic spokesperson Danielle Ghiglieri stated that the company has received confirmation from the U.S. government that Mythos 5, their top cybersecurity model, can be redeployed to a limited group of cyber defenders and infrastructure providers. She mentioned, “We are working to provision the approved set of providers and restore their access to Mythos 5 as quickly as possible. We are pleased to see this progress and continue to work with the government to expand access to Mythos 5 and make Fable 5 available for general use again.”
The U.S. government has not lifted the export control directive imposed on Anthropic two weeks ago, which prohibited any foreign national from accessing either model, including Anthropic’s own employees. Instead, the government has made an exception for Mythos 5, allowing a select group of organizations access, similar to the arrangement made for OpenAI’s GPT-5.6, which was announced earlier. Under this exception, Anthropic employees who are not U.S. nationals, as well as members of the approved organizations, are now permitted to access Mythos 5.
“These efforts have yielded significant progress,” Lutnick wrote. “In addition, Anthropic has committed to work with the U.S. government on protocols and standards for [Mythos-class models]. In light of this progress … I have determined that appropriate safeguards are in place to permit certain trusted partners to access the Claude Mythos 5 Model.”
There has been increasing pressure on the Trump administration to shift away from the recent case-by-case regulatory approach, especially as Anthropic’s competitors have been enhancing their cybersecurity models and even surpassing it in some benchmarks. Concerns have also arisen within the U.S. AI industry regarding potential advancements in AI by China while leading U.S. labs remain sidelined. Additionally, key U.S. government departments, such as the National Security Agency, had lost access to Mythos 5.
Now, Anthropic has a deal similar to OpenAI’s: releasing the model in a limited preview for approved organizations, including trusted enterprises and the U.S. government. Both AI labs are hopeful for broader availability soon, both for enterprise deals and public access (like with Anthropic’s Fable 5), but the final decision rests with the Trump administration. While many have called for AI regulation, including from within the AI labs, there is a growing consensus among some tech leaders that the past two weeks have not been the ideal way to approach it.
OpenAI expressed in their GPT-5.6 blog post, “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. We are taking this short-term step because we believe it is the strongest path to broader availability in the coming weeks, while we work with the Administration to develop the cyber Executive Order framework and a repeatable process for future model releases.”
Lutnick noted in his letter to Anthropic, “All other requirements of the June 12 letter remain in effect until further notice.” He also stated, “I reserve the right to reevaluate and adjust the scope of license requirements on [Mythos 5 and Fable 5], should circumstances change.”
