The Anthropic Mythos 5 is back

After rollercoaster negotiations with the Trump administration that lasted for two weeks, Anthropic’s Mythos 5 is back in action – maybe, just maybe, for a select group, according to a letter from the government to Anthropic that was seen by Seaside. Myth 5, however – a version of the Mythos-class focused on the group – appears to be still in development, with no apparent timeline for the alliance.

The letter, dated June 26th and sent by the Secretary of Commerce Howard Lutnick to the co-founder of Anthropic Tom Brown, who was leading the discussion recently, says that there has been a “renewal of the licensing requirements” in light of the fact that Anthropic has recently “worked with the US government to deal with the risks” associated with Mythos 5 and Fable 5.

Anthropic spokesperson Danielle Ghiglieri shared this Seaside in a statement the company “received information from the US government that Mythos 5, our most powerful cybersecurity model, can be redeployed to a small group of cyber defenders and infrastructure providers.” He added, “We are working to provide a valid group of service providers and restore their access to Mythos 5 as quickly as possible. We are happy to see this progress and continue to work with the government to increase access to Mythos 5 and make Fable 5 available for regular use.”

The US government has not solved the problem export control instructions it reached Anthropic two weeks ago, who banned aliens from accessing any species (including its Anthropic employees). Instead, the government chose to abandon Mythos 5, allowing a select group of organizations to access the same channels. OpenAI’s GPT-5.6it was announced earlier today. With this exception, Anthropic employees who are not US citizens and members of the official organizations that are not US citizens are all green to get Mythos 5, according to the letter.

“This has been very helpful,” Lutnick wrote. In light of this situation… I have determined that appropriate safeguards are in place to allow other trusted individuals access to the Claude Mythos Specimen 5.”

Pressure he has been building on the Trump administration to change the type of legislation he received recently, especially if Anthropic’s examples of cybersecurity competition were going well – and even rising ahead of other cybersecurity indicators. Pressure was building in the US AI industry, especially over fears about the AI ​​advances China could make while top US AI labs were sidelined. In addition, there was also the fact that the top departments of the US government, such as the National Security Agency, were involved he lost his chance in Mythos 5.

Now, Anthropic has the same commitment that OpenAI has: Release the model in limited previews, only for official organizations, such as trusted businesses and the US government itself. All the AI ​​labs are hoping that more availability will come soon, both for business purposes and for people to reach (like with Anthropic’s Fable 5), but ultimately it will be up to the Trump administration. While many have called for AI control, including AI labs themselves, a current agreement among other technology leaders that the last two weeks have not been the way to do it.

Instead, OpenAI wrote in the company’s GPT-5.6 blog post“We do not believe that such government access measures should remain unchanged for the long term. It preserves the best tools from users, developers, businesses, cyber defenders, and partners around the world who need them. We are doing this for a short period of time because we believe it is the most powerful way to achieve greater availability in the coming weeks, while we work with the Government to develop a cyber Executive Order plan and repeat the model.”

Lutnick wrote in his letter to Anthropic, “All other requirements of the June 12 letter remain in effect until further notice.” He added, “I reserve the right to review and change the license amount for (Mythos 5 and Fable 5), if the situation changes.”