Red Hat Quay CVE Lets Attackers Overwrite All Private Containers
A critical vulnerability in Red Hat Quay container registry allows attackers to read and overwrite private container images across every repository. This flaw, tracked as CVE-2026-123710, carries a CVSS score of 9.8 and has existed since Quay’s first public commit. Red Hat disclosed the vulnerability this week along with a second critical issue that lets attackers steal tenant robot account tokens.
What Happened: Critical Red Hat Quay Container Registry Vulnerability Disclosed
Red Hat disclosed two critical vulnerabilities in its Quay container registry platform on November 13, 2026. The first flaw, CVE-2026-123710, enables an attacker with network access to a Quay registry to add their own trusted signing key. This key has no expiry date and grants full access to every private repository.
Furthermore, Quay enables anonymous access by default. While anonymous access remains enabled, an attacker needs no account at all. Even when administrators disable anonymous access, anyone with a valid account can still exploit the flaw.
A second critical flaw, CVE-2026-110697 (CVSS score: 9.9), was also disclosed at the same time. This vulnerability lets a Quay user who can create a namespace steal other tenants’ robot account tokens. These tokens can then be used to read, push, or overwrite images belonging to other tenants.
Technical Details of the CVE-2026-123710 Authentication Bypass
Quay uses trusted keys to sign the access tokens that allow image pushes and pulls. The vulnerability exists in the key server’s validation logic.
Specifically, Quay’s key server checks that a request to replace a key carries a token signed by a trusted key with the expected audience and issuer. However, it does not check whether the token was actually issued for key rotation. Consequently, an anonymous token from Quay’s /v2/auth endpoint passes validation. Quay signs this token with a trusted key, and the caller chooses its own audience.
The attack works as follows:
- The attacker requests an anonymous token from the
/v2/authendpoint. - The attacker uses this token to add a new trusted signing key via the
/keysendpoint. - The planted key has no expiry date and survives registry restarts.
- The attacker signs new tokens with this key to access any repository.
The attacker can also delete existing keys that running Quay instances use. This causes normal pushes and pulls to fail with authentication errors, creating a denial of service condition.
Secondary CVE-2026-110697: SQL LIKE Wildcard Token Theft
The second flaw exists in the code that lists robot accounts. The code builds an SQL LIKE pattern from the namespace name without escaping wildcard characters. In SQL LIKE, an underscore matches any single character.
Therefore, a namespace with an underscore in the right place can match another tenant’s namespace and return its robot accounts with their decrypted tokens. The attacker needs no elevated privileges, and the victim does not have to take any action.
Deleting the attacker’s namespace also deletes the corresponding robot accounts, which can disrupt systems that rely on them.
Business and Operational Impact
The impact of these vulnerabilities is severe and widespread. Red Hat lists every version of Quay 3 and Quay 5 as affected by both flaws.
Key impacts include:
- Data exposure: Attackers can retrieve private container images and any secrets or credentials stored within them.
- Supply chain compromise: Attackers can push or overwrite images, and these changes propagate to every system that pulls them.
- CI/CD pipeline injection: Where Quay is the integrated registry for OpenShift, stolen robot credentials can inject malicious images into CI/CD pipelines.
- Denial of service: Deleting trusted keys breaks normal authentication for pushes and pulls.
- Persistence: The planted key survives restarts and remains effective after password resets or robot token changes.
Any Quay instance that untrusted clients can reach is exposed, including internal registries.
Mitigation and Recommendations
Red Hat has released fixed builds for eight Quay 3 branches. Quay.io is already patched. However, the upstream CVE entry still showed every Quay package as affected with no fix at the time of writing, so administrators should verify their specific version against Red Hat’s advisories.
Immediate Actions for Defenders
- At the proxy or load balancer, block the
/keysendpoint for everything except trusted internal components. - Search audit logs for
service_key_rotateandservice_key_deleteevents whereexpiration_dateis null. - Remove any approved service keys you do not recognize.
- Install the fixed release as soon as possible.
- For CVE-2026-110697, limit who can create organizations or namespaces containing an underscore.
- Rotate robot tokens for any namespace that matches an underscore namespace.
- Watch for pulls or pushes that do not fit normal robot use patterns.
Important Limitations
Turning off anonymous access does not fully protect against CVE-2026-123710. Anyone with a valid account can still exploit the flaw. Also, deleting an unauthorized key is not a lasting fix because an attacker can add a new key the same way until the patched release is running.
Bottom line: Every Quay instance reachable by untrusted clients is at risk. Patch immediately, block the /keys endpoint at the network edge, and audit for unauthorized keys.
Incident Summary
| CVE ID / Incident: | CVE-2026-123710 (CVSS 9.8) and CVE-2026-110697 (CVSS 9.9) |
| Affected Systems: | All versions of Red Hat Quay 3 and Quay 5; standalone installs, Quay Operator on OpenShift, and OpenShift mirror registry |
| Disclosure Date: | November 13, 2026 |
| Patch Status: | Fixed builds released for eight Quay 3 branches; Quay.io patched |
| Attack Vector: | Network; anonymous access enabled by default, valid account required if anonymous disabled |
| Discovered By: | Dave O’Connor (Red Hat) |
References
- The Hacker News, “Critical Red Hat Quay Flaw Lets Attackers Read and Overwrite All Private Container Images in a Registry,” November 13, 2026, https://thehackernews.com/2026/11/critical-red-hat-quay-flaw-lets.html.
- Red Hat Security Advisory (CVE-2026-123710), accessed November 14, 2026.
- Red Hat Security Advisory (CVE-2026-110697), accessed November 14, 2026.