CVE-2026-73122 Overview
CVE-2026-73122 is a privilege management flaw [CWE-269] in the multicloud-operators-channel component of Red Hat Advanced Cluster Management (RHACM). A compromised agent running on a managed cluster can read all Secrets and ConfigMaps within any Channel namespace on the hub cluster. This cross-tenant access can expose credentials for other tenants' Git and Helm repositories used by RHACM subscriptions. The result is significant information disclosure across tenant boundaries in a multi-tenant hub deployment.
Critical Impact
A single compromised managed-cluster agent can harvest Git and Helm repository credentials belonging to unrelated tenants sharing the RHACM hub, enabling downstream supply-chain compromise of application delivery pipelines.
Affected Products
- Red Hat Advanced Cluster Management for Kubernetes (RHACM)
- multicloud-operators-channel component on the RHACM hub
- Multi-tenant RHACM deployments where Channel namespaces host Git or Helm subscription credentials
Discovery Timeline
- 2026-08-12 - CVE-2026-73122 published to the National Vulnerability Database
- 2026-08-12 - Last updated in NVD database
Technical Details for CVE-2026-73122
Vulnerability Analysis
RHACM uses the multicloud-operators-channel controller on the hub to distribute application manifests sourced from Git and Helm channels to managed clusters. Each Channel object lives in a namespace and typically references Secret and ConfigMap resources that hold repository authentication material. The controller grants managed-cluster agents access rights broader than the specific Channel they subscribe to. As a consequence, an agent on any managed cluster can enumerate and read every Secret and ConfigMap in any Channel namespace on the hub. The flaw is a scope containment failure at the Kubernetes RBAC layer, aligning with CWE-269 (Improper Privilege Management).
Root Cause
The root cause is over-permissive role bindings tied to the managed-cluster agent's service account on the hub. Instead of scoping read access to the resources required for a cluster's assigned subscriptions, the bindings extend to all Channel namespaces. Kubernetes then evaluates the request as authorized, even when the requesting agent is not a legitimate consumer of that Channel.
Attack Vector
An attacker who compromises any managed-cluster agent, for example through a workload escape or stolen kubeconfig, reuses the agent's credentials against the hub API server. Using standard kubectl get secret or kubectl get configmap requests against Channel namespaces, the attacker exfiltrates Git tokens, Helm registry credentials, and related configuration. The scope-changed CVSS vector reflects that the impact crosses the tenant boundary from the managed cluster into the hub's shared namespaces.
No verified public exploit code is available. See the Red Hat CVE-2026-73122 Advisory and Red Hat Bugzilla Report #2514230 for authoritative technical details.
Detection Methods for CVE-2026-73122
Indicators of Compromise
- Kubernetes audit log entries showing get or list verbs on secrets or configmaps in Channel namespaces where the requesting user is a managed-cluster agent service account not owning that Channel.
- Access to hub Channel namespaces originating from managed-cluster agent identities in short bursts or enumeration patterns.
- Unexpected outbound authentication attempts against Git or Helm repositories using credentials that should be scoped to a different tenant.
Detection Strategies
- Enable and centralize Kubernetes API audit logs on the RHACM hub, with a rule set that flags reads of Secret and ConfigMap objects in namespaces containing Channel custom resources.
- Baseline which managed-cluster agent identities legitimately access each Channel namespace, then alert on any deviation.
- Correlate hub audit events with managed-cluster workload activity to identify agents whose behavior changes after a suspected compromise.
Monitoring Recommendations
- Forward hub kube-apiserver audit logs and RHACM controller logs to a central analytics platform for retention and querying.
- Monitor Git and Helm repository access logs for authentications from unexpected source clusters or IP ranges.
- Track creation and modification of RoleBindings and ClusterRoleBindings tied to open-cluster-management-agent service accounts.
How to Mitigate CVE-2026-73122
Immediate Actions Required
- Apply the RHACM security update referenced in the Red Hat CVE-2026-73122 Advisory as soon as it is available for your subscribed release.
- Rotate all Git and Helm repository credentials stored in Channel namespaces on the hub, treating them as potentially disclosed in multi-tenant environments.
- Audit historical Kubernetes API access on the hub for reads of Channel namespace Secrets and ConfigMaps by managed-cluster agents.
Patch Information
Red Hat tracks remediation for CVE-2026-73122 under the advisory published at access.redhat.com and the associated Bugzilla report #2514230. Administrators should consult the advisory for the fixed multicloud-operators-channel component version corresponding to their RHACM release and apply it through the standard operator update channel.
Workarounds
- Segregate tenants across separate RHACM hubs where strict tenant isolation is required, until the patched component is deployed.
- Tighten RBAC on Channel namespaces by removing wildcard read access and restricting managed-cluster agent service accounts to only the Channels they consume.
- Store repository credentials in an external secrets manager and inject them at subscription time rather than persisting them in hub Channel namespaces.
Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.
