Secure YAML, Insecure Clusters: Breaking Kubernetes Without Exploits
How can Kubernetes clusters with clean scan results still be compromised, with no kernel exploits, no zero-days, and no exotic techniques?
Abstract
Kubernetes security is increasingly reduced to static configuration reviews and misconfiguration scanners. While these tools provide value, they are often treated as authoritative indicators of real security posture.
In this talk, we will show how Kubernetes clusters with clean scan results can still be compromised through realistic attack paths that require no kernel exploits, no zero-days, and no exotic techniques.
Drawing on academic research, real incidents, and offensive security experience, we demonstrate how attackers chain legitimate Kubernetes features into high-impact compromises.