Two Critical Red Hat Quay Flaws Put Private Container Images at Risk
Red Hat has released fixes for two critical flaws in Quay, its container registry. The first, CVE-2026-123710 (CVSS 9.8), lets an attacker who can reach a Quay registry read and overwrite private images in every repository. Quay allows anonymous access by default, so no account is needed while that setting is on. With it off, any valid account can still exploit the flaw. Red Hat says the attacker can plant their own trusted signing key with no expiry date, and that key survives restarts until someone removes it.
With that key, an attacker can pull private images and any secrets or credentials stored inside them. They can also push or overwrite images, and the change spreads to every system that later pulls them. The flaw can also be used to delete the keys running instances rely on, which causes pushes and pulls to fail with authentication errors. The cause is that Quay's key server does not check whether a token was issued for key rotation.
A second flaw, CVE-2026-110697 (CVSS 9.9), was disclosed at the same time. It lets a user who can create a namespace steal other tenants' robot account tokens, then read, push or overwrite their images. The victim does not need to do anything. Where Quay is the integrated registry for OpenShift, Red Hat says stolen robot credentials could be used to inject malicious images into CI/CD pipelines. Red Hat has released fixed builds for eight Quay 3 branches.