Listen to this Post
How CVE-2026-44938 Works
CVE-2026-44938 is a high-severity vulnerability identified in Rancher Fleet’s agent-side deployer. The core of the issue lies in how Fleet processes the `namespaceLabels` field, which can be defined in a `fleet.yaml` file or within BundleDeployment.spec.options.namespaceLabels.
Kubernetes clusters use Pod Security Standards (PSS) to enforce security policies on pods. These standards are applied to namespaces through specific labels, such as pod-security.kubernetes.io/enforce: baseline. The vulnerability stems from Fleet’s failure to filter these critical, security-sensitive labels when applying them to a target namespace.
An attacker with `git push` access to a repository monitored by Fleet could exploit this. By crafting a malicious `fleet.yaml` or `BundleDeployment` with a `namespaceLabels` section, they could overwrite the PSS enforcement labels on a target namespace.
For example, an attacker could change the enforcement level from `restricted` to `baseline` or privileged, thereby weakening the admission controls that govern what pods can run in that namespace. This would allow them to deploy workloads that would normally be blocked by the cluster’s security policies, such as privileged containers.
The final impact on confidentiality, integrity, and availability is contingent on the specific permissions of any credentials that might be leaked or abused as a result of the weakened security posture. The Fleet team recommends reviewing systems for potentially leaked credentials and replacing any that may have been compromised. This attack falls under the MITRE ATT&CK technique “Disable or Modify Tools”.
DailyCVE Form
Platform: ……. Rancher Fleet
Version: …….. <0.12.15, <0.13.11, <0.14.6, <0.15.2
Vulnerability: …. PSS Bypass via namespaceLabels
Severity: ……. High
date: ………. July 1, 2026
Prediction: …… July 1, 2026
What Undercode Say
This section provides technical analytics and commands related to the vulnerability.
Identifying Vulnerable Versions
To check your Fleet version, you can query the Fleet controller image tag:
kubectl get deployment -n fleet-system fleet-controller -o jsonpath='{.spec.template.spec.containers[bash].image}'
Simulating the Attack (Conceptual)
An attacker with push access could create a `fleet.yaml` like this:
namespace: target-namespace namespaceLabels: pod-security.kubernetes.io/enforce: privileged
Exploit
An attacker exploits this vulnerability by pushing a malicious configuration to a Fleet-monitored Git repository. The vulnerable Fleet agent then deploys the change, applying the attacker’s labels to the namespace, thereby overwriting the cluster administrator’s PSS settings. This action weakens the security posture of the namespace, allowing the attacker to deploy pods that would otherwise be blocked by policies.
Protection
The primary protection against this vulnerability is to upgrade to a patched version of Fleet. The fixed versions are v0.15.2, v0.14.6, v0.13.11, and v0.12.15. These versions include a fix that filters out labels with the `pod-security.kubernetes.io/` prefix when applying namespaceLabels.
If an immediate upgrade is not possible, two workarounds are available:
1. Deploy NeuVector (Primary Workaround): Deploy NeuVector (SUSE Security) and configure an admission control rule to deny pods running as privileged. This works because NeuVector evaluates pod specifications independently of Kubernetes PSS namespace labels.
2. Restrict Repository Access (Secondary Workaround): Restrict `git push` access to Fleet-monitored repositories, especially in multi-tenant setups. This reduces the attack surface but does not fully close the vulnerability.
Impact
A successful exploit allows an attacker to bypass Pod Security Standards, a core Kubernetes admission control mechanism. By lowering the security level of a namespace, the attacker can deploy privileged or otherwise restricted workloads. The ultimate impact on the cluster’s Confidentiality, Integrity, and Availability (CIA) triad is variable and depends on the privileges of the deployed workloads and any subsequent actions the attacker takes.
🎯Let’s Practice Exploiting & Learn Patching For Free:
🎓 Live Courses & Certifications:
Join Undercode Academy for Verified Certifications
🚀 Request a Custom Project:
Secure, high-velocity infrastructure and disruptive technological engineering. Contact our engineering team for high-tier development and proprietary systems:
[email protected]
💎 Smart Architecture | 🛡️ Secure by Design | ⭐ Trusted by Thousands
Sources:
Reported By: github.com
Extra Source Hub:
Undercode
