/

CVE-2026-49823

Severity:

HIGH RISK

(7.7)

Authorization flaw (CWE-284) in Fission function specs

CVE-2026-49823

Deferred

Vulnerability type

Category:

Authorization and Access Control

CWE:

CWE-284, CWE-863

Impact:

Privilege Escalation / Cross-tenant Access

Severity:

HIGH RISK

(7.7)

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

An authorization flaw in Fission function specs allowed the PackageRef namespace to bypass admission webhook validation prior to version 1.24.0.

Under certain deployment configurations, this could let a tenant reference packages in other namespaces, potentially leading to privilege escalation or cross-namespace access.

Affected context

This issue affects Fission deployments on Kubernetes where:

  • Multiple tenants or teams share a cluster using namespace boundaries
  • Function specs are created or updated by users with limited intended scope
  • Admission webhooks are relied upon to enforce namespace-level access controls

In these environments, missing namespace validation on PackageRef can undermine the isolation assumed between namespaces.

Exploit preconditions

  • Attack surface: Fission function spec creation or modification (API/CLI/CI pipelines).
  • Required access: Ability to create or update Fission Functions in at least one namespace.
  • Trigger point: Submitting a Function spec whose PackageRef.Namespace points to a different, otherwise restricted namespace.
  • Why exploitation is possible: The admission webhook validates Secret and ConfigMap namespaces but does not enforce namespace checks on PackageRef, allowing cross-namespace references where they should be disallowed.

Why this happens

The root cause is incomplete authorization and namespace validation logic in the admission webhook responsible for Fission Function specs.

While Secret and ConfigMap references were checked against the caller's namespace, PackageRef.Namespace was left unchecked, creating an inconsistent enforcement pattern.

This maps to CWE-284 (Improper Access Control) and CWE-863 (Incorrect Authorization), where a security control (namespace-based isolation) is only partially applied, permitting resources to be bound across privilege boundaries.

Potential impact

  • A function in one namespace may reference and execute code packages from another namespace.
  • Tenants could gain indirect access to packages they are not intended to use.
  • Cross-namespace privilege escalation could occur if sensitive or more privileged packages are referenced.
  • Breakdown of assumed multi-tenant isolation and potential data or code exposure.

General mitigation guidance

Mitigation should focus on enforcing consistent, least-privilege access control across all referenced resources in function specifications.

Recommended patterns include:

  • Applying uniform namespace validation to all reference types (Secrets, ConfigMaps, Packages, and similar resources).
  • Centralizing authorization checks so that any new reference fields must pass the same policy gates.
  • Designing admission and policy logic to default to denial when namespace or ownership information is missing or ambiguous.
  • Using Kubernetes RBAC and network policies to reinforce namespace isolation so that unauthorized cross-namespace bindings are not useful even if misconfigured.
  • Treating function specs as untrusted input and validating every external reference against the caller's identity and allowed scope.

How teams detect issues like this

Detection can rely on both configuration review and runtime policy enforcement:

  • Static analysis or policy-as-code (e.g., OPA/Gatekeeper, Kyverno) to flag Function specs whose PackageRef.Namespace differs from the Function's namespace or violates tenancy rules.
  • Admission controller logs and audit events to identify cross-namespace PackageRef usage and anomalous references.
  • Periodic cluster configuration audits to ensure all reference fields in Fission resources are covered by consistent validation policies.
  • Automated CI/CD checks that reject manifests where function references span namespaces outside of approved patterns.

Sources

SourceLink
github.comgithub.com/fission/fission/pull/3389
github.comgithub.com/fission/fission/releases/tag/v1.24.0
github.comgithub.com/fission/fission/security/advisories/GHSA-3r8v-2xmj-5c39

Frequently asked

questions

It is exploitable through the Fission or Kubernetes API surfaces that accept Function specs, which may be exposed internally or externally depending on how the cluster is configured. An attacker must be able to submit or modify Function resources.

The attacker needs permissions to create or update Fission Function resources in at least one namespace. They do not need direct access to the target namespace whose packages they are attempting to reference.

The safest pattern is to enforce uniform authorization and namespace validation on all resource references in Function specs, treating each reference field as a potential cross-boundary link and denying it unless explicitly allowed by policy.

Prevent similar issues by using centralized, policy-driven authorization (e.g., admission policies), applying deny-by-default semantics to new fields, and incorporating security reviews and automated tests that verify namespace isolation for every new reference type introduced into the API.

Find out what's already
exploitable in your codebase.