Rubrik Launches Code Guardian to Defend Enterprise Code Against Frontier AI Threats

Rubrik has expanded Project Hourglass with Rubrik Code Guardian, a new security service powered by Anthropic's Claude Mythos 5 that red-teams enterprise codebases to uncover sophisticated, multi-step vulnerabilities exploitable by frontier AI models. The service operates on isolated, air-gapped copies of customer repositories, using Claude Mythos 5 within a custom security harness to discover attack chains that span files, services, identity permissions, and cloud boundaries—then validates each finding for actual exploitability and ranks it by business impact.