U.S. Anthropic revealed on its official X account that it had received notice from the U.S. government allowing access to its AI model “Claude Mythos 5” to resume, limited to select U.S. organizations that operate and defend critical infrastructure, as of June 26 (local time). “Claude Fable 5” is not included.
Mythos 5, which the company describes as its “most powerful cybersecurity model,” is set to partially resume following the blanket suspension imposed by export-control measures on June 12.
“The model we can call our most capable in cybersecurity and biological research”
Anthropic said it would restore access for approved recipients as quickly as possible and would continue discussions with the government toward expanding access to Mythos 5 and resuming the availability of the consumer-facing model “Claude Fable 5.”
Fable is positioned as “next-generation intelligence for tackling the most challenging knowledge-work and programming problems”
U.S. CNN and other outlets reported that the measure was based on a letter dated the 26th sent by Commerce Secretary Howard Lutnick to Anthropic. Lutnick said that, following the export restrictions, the company had been “working to address the risks relating to the targeted models,” and, taking note of that progress, had “determined that appropriate safeguards are in place and that access to the Claude Mythos 5 model should be allowed for a select group of trusted partners.”
Regarding the organizations eligible for redeployment, Anthropic’s statement merely referred to them as “a range of U.S. organizations that operate and defend critical infrastructure” and “a small group of cyber defenders and infrastructure providers,” without specifying an exact number. According to several foreign media outlets, around 100 organizations, including government agencies and private companies, are expected to be included.
