A group of cybersecurity experts is calling on the White House to lift export controls on Anthropic's AI models Fable and Mythos. The restrictions, the experts argue, hinder cybersecurity research and leave defenders at a disadvantage against adversaries. The appeal was published this week by Crypto Briefing, highlighting a growing friction between national security policy and the practical needs of the security community.
Why the controls are under fire
The experts contend that keeping Fable and Mythos off the export list prevents security teams from using state-of-the-art AI to detect and respond to threats. Without access, defenders rely on older tools, while attackers — who operate outside such restrictions — may already be leveraging similar capabilities. The letter frames the policy as counterproductive, weakening the very defenses export controls are meant to protect.
Fable and Mythos by Anthropic
Anthropic, a leading AI safety company, developed Fable and Mythos as advanced language models. Both are currently subject to U.S. export controls that limit their distribution abroad. The experts' request does not challenge the rationale behind export controls broadly, but argues these specific models pose no national security risk and should be freely available for defensive cybersecurity work.
What happens next
Crypto Briefing obtained the letter but did not report any response from the White House as of publication. The administration has been reviewing its AI export framework since early 2026, and this appeal adds pressure to carve out exemptions for cybersecurity applications. For now, Fable and Mythos remain restricted — and the defenders who asked for access are still waiting.
