TECH & AI
What all of these new partners have in common, Anthropic says, is that a major attack on any of these would have“ catastrophic” results.
“ For most partners, we estimate that a major attack could affect more than 100 million people, with important ramifications for both global and national security,” the AI giant elaborates.
An aggressive plan for Mythos-safe security The cyber industry, as it usually is, was swift in its response, with a community of security leaders publishing a draft paper with involvement from the Cloud
Security Alliance( CSA) alongside SANS Institute and OWASP contributors titled: The‘ AI Vulnerability Storm’: Building a‘ Mythosready’ Security Program.“ Even with active human defenders and alerts, the speed asymmetry is what should keep you up,” says Rob T. Lee, Chief AI Officer and Chief of Research at SANS Institute, who was one of the authors of the paper.“ The window between vulnerability discovery and weaponisation has collapsed into hours.” All it takes to prove Rob’ s point is to look up the Zero Day Clock online, which shows the time it takes between confirming a vulnerability and active exploitation.
126 July 2026