A group of influential cybersecurity leaders, including CISOs, security researchers, and executives from Adobe, Zoom, and Sophos, are urging the Trump administration to revisit the restrictions placed on the Mythos and Fable models.
In an open letter titled “On Transparent AI Cyber Protections,” they argue that this decision will deprive defenders of the best models available, create market uncertainty, and jeopardize American leadership in AI, without any real risk justifying such a move.
Led by Alex Stamos, the former head of security at Facebook and current Director of Products at Corridor, the initiative already counts more than 40 signatories.
Among them are Katie Moussouris (CEO of Luta Security), Rachel Tobac (CEO of SocialProof Security), Chris Wysopal (co-founder of Veracode), Joe Levy (CEO of Sophos), and Aaron Grattafiori (security researcher at Nvidia).
An Initiative Led by Industry Figures
According to Alex Stamos, the security capability of the Fable 5 model that appears to have alarmed the White House lies in its ability to forge “proofs of concept” for vulnerabilities.
If these proofs of concept can serve as an action plan for attackers seeking entry into a system, they also enable security teams to understand how to protect their infrastructure.
Alex Stamos notes that only the Mythos 5 and Mythos Preview models, which were reserved before June 12 for the hand-picked members of the Project Glasswing initiative, are capable of turning these proofs of concept into fully autonomous attack chains.
He adds that it would be impossible to hand over the entire Linux kernel to Fable and expect it to identify every security bug.
The letter’s authors emphasize that this analytical capability can also be replicated with other tools, such as OpenAI’s GPT-5.5, Anthropic’s Opus and Sonnet models, as well as Chinese models like Kimi 2.7.
Alex Stamos insists that Chinese open-source models are not far behind Fable 5 in their ability to analyze security flaws.
According to him, restricting the best American capabilities at the precise moment when China is using and storing these vulnerabilities is a dangerous and ill-advised decision, noting that teams are engaged in a race to fix these bugs as quickly as possible.
The letter remains open to new signatures from industry professionals for the moment.
