/

CVE-2026-55225

Severity:

HIGH RISK

(8.0)

Kubernetes Privilege Escalation (CWE-269) in Strimzi

CVE-2026-55225

Awaiting Analysis

Vulnerability type

Category:

Privilege Escalation

CWE:

CWE-269, CWE-441, CWE-250

Impact:

Cross-Namespace Secret Access and Modification

Severity:

HIGH RISK

(8.0)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

A privilege-management flaw in Strimzi can cause the Cluster Operator to grant broad Secret permissions in a target namespace to an Entity Operator service account controlled from an attacker's namespace.

An attacker who can create a Kafka custom resource may then obtain access to Secrets across namespaces where the Cluster Operator is authorized, allowing unauthorized reading or modification of sensitive data.

Affected context

The issue occurs in Strimzi deployments where users can create or modify Kafka custom resources and the Cluster Operator processes the Kafka.spec.entityOperator.watchedNamespace field. Risk increases when the Cluster Operator has permissions across multiple namespaces and the resulting service account can receive tokens. The vulnerable behavior can cross the intended STRIMZI_NAMESPACE boundary.

Exploit preconditions

  • Attack surface: Kubernetes or OpenShift API through Kafka custom resources
  • Required access: Permission to create a Kafka custom resource in an attacker-controlled namespace
  • Trigger point: Cluster Operator reconciliation of the Entity Operator namespace configuration
  • Required deployment condition: The Cluster Operator must be able to create roles and role bindings in the selected target namespace
  • Why exploitation is possible: The operator creates full Secret CRUD permissions in the target namespace and binds them to a service account associated with the attacker's namespace

Why this happens

The underlying flaw is improper privilege management across namespace boundaries. Strimzi accepts a user-controlled watched-namespace value and uses it to create authorization resources without sufficiently constraining the relationship between the requested namespace, the service account, and the operator's intended scope.

This maps primarily to CWE-269: Improper Privilege Management, with related characteristics of CWE-441: Unintended Proxy or Intermediary ('Confused Deputy') and CWE-250: Execution with Unnecessary Privileges. The Cluster Operator acts with elevated authority and unintentionally delegates that authority to an attacker-controlled service account.

Potential impact

  • An attacker may read sensitive Kubernetes Secrets from authorized target namespaces.
  • An attacker could modify or delete Secrets, disrupting workloads or changing application configuration.
  • Credentials, tokens, or connection information may be exposed.
  • Compromised Secrets could enable further access to applications, infrastructure, or data.
  • Namespace isolation and organizational access boundaries could be weakened.

General mitigation guidance

Mitigation generally focuses on constraining operator-delegated permissions and treating namespace selection as an authorization decision rather than ordinary configuration.

Common defensive patterns include:

  • Restricting Kafka custom-resource creation and modification to trusted identities.
  • Allowing watched namespaces only through an explicit, policy-controlled allowlist.
  • Scoping service-account permissions to the minimum resources and namespaces required.
  • Avoiding full Secret create, read, update, and delete permissions when narrower access is sufficient.
  • Separating operator privileges from tenant-controlled service accounts.
  • Applying Kubernetes admission policies to reject cross-namespace configurations that violate tenancy boundaries.
  • Auditing and periodically reviewing ClusterRole, Role, RoleBinding, and service-account relationships.
  • Using least-privilege tokens and limiting token creation capabilities.

How teams detect issues like this

Detection can combine configuration review, authorization analysis, and runtime monitoring.

Useful approaches include:

  • Static and policy analysis for user-controlled namespace fields that influence Role or RoleBinding creation.
  • Code review checks for operators that copy privileges across namespaces or bind elevated permissions to tenant-controlled service accounts.
  • Admission policies that flag or deny watched namespaces outside an approved scope.
  • Auditing Kubernetes API events for unexpected Role, RoleBinding, Secret, or service-account token activity.
  • Reviewing effective permissions with authorization tools and comparing them against intended namespace boundaries.
  • Monitoring for service accounts reading or modifying Secrets outside their normal workload namespace.

Sources

SourceLink
github.comgithub.com/strimzi/strimzi-kafka-operator/commit/b3bfeffcc30754c3e62e6f4afd8a76942cac440d
github.comgithub.com/strimzi/strimzi-kafka-operator/commit/f6c5207ba7ec89b46ce6720b8638d647e556a261
github.comgithub.com/strimzi/strimzi-kafka-operator/pull/12844
github.comgithub.com/strimzi/strimzi-kafka-operator/releases/tag/1.0.1
github.comgithub.com/strimzi/strimzi-kafka-operator/releases/tag/1.1.0
github.comgithub.com/strimzi/strimzi-kafka-operator/security/advisories/GHSA-mw9r-p8xp-wx96
access.redhat.comaccess.redhat.com/errata/RHSA-2026:54435
access.redhat.comaccess.redhat.com/security/cve/CVE-2026-55225
bugzilla.redhat.combugzilla.redhat.com/show_bug.cgi?id=2490275
security.access.redhat.comsecurity.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-55225.json

Frequently asked

questions

It can be reached through the Kubernetes or OpenShift API, but exploitation requires authenticated access that permits creation or modification of the relevant Kafka custom resource.

Not necessarily. The security boundary is weakened because a service account associated with the attacker's namespace may receive Secret permissions in a different target namespace, provided the Cluster Operator is authorized there.

Constrain namespace selection through explicit authorization policy, ensure operator-generated permissions remain within the intended scope, and apply least privilege to service accounts and token operations.

Treat custom-resource fields that influence RBAC as security-sensitive, enforce admission controls, review generated authorization objects, and continuously audit effective permissions across namespaces.

Find out what's already
exploitable in your codebase.