Skip to main content
Vulnerability Database/CVE-2026-94611

CVE-2026-94611: authentik Information Disclosure Vulnerability

CVE-2026-94611 is an information disclosure vulnerability in authentik open-source identity provider that exposes stored credentials to unauthorized viewers. This post covers technical details, affected versions, and remediation.

Published:

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

CVE-2026-94611 Overview

authentik, an open-source identity provider, exposes stored credentials through its API serializers to accounts that hold only view permissions on affected configurations. The flaw affects deployments where view permission is granted to accounts not intended to read secrets. Impacted configurations include one-time code delivery by mail or SMS, outbound provisioning targets, device trust integrations, identity sources, the Kubernetes outpost integration, applications using a client or shared secret, and applications using a proxy provider. The issue is tracked as [CWE-200] information exposure and is fixed in authentik 2026.2.7, 2026.5.7, and 2026.8.2.

Critical Impact

Low-privileged authenticated users can read stored credentials for identity sources, outbound provisioning connectors, Kubernetes outposts, and application client secrets, enabling lateral movement and downstream compromise of integrated systems.

Affected Products

  • authentik versions prior to 2026.2.7
  • authentik versions prior to 2026.5.7
  • authentik versions prior to 2026.8.2

Discovery Timeline

  • 2026-09-24 - CVE-2026-94611 published to NVD
  • 2026-09-24 - Last updated in NVD database
  • Vendor advisory published as GHSA-m9h4-7j9c-55x9

Technical Details for CVE-2026-94611

Vulnerability Analysis

The vulnerability resides in authentik's Django REST Framework serializers. The to_representation method returned every field defined in Meta.fields, including fields that store secrets such as credentials, client_secret, and connector tokens. Access was gated only by Django's view_* permission on the model, which many deployments grant broadly for operational visibility.

An authenticated user with view permission on an affected object could issue a GET request against the corresponding API endpoint and receive the raw secret value in the JSON response. The attacker did not need change_* permission on the object nor any explicit secret-read entitlement. Affected object types span identity sources, outbound provisioning providers (for example, Google Workspace), device trust connectors, the Kubernetes outpost configuration, OAuth2/OIDC applications with client secrets, and proxy provider applications.

Root Cause

The root cause is missing field-level authorization in the base serializer. Permission enforcement was applied at the view level for read access, but the serializer did not distinguish between fields safe for viewers and fields that should remain restricted to accounts able to modify the object. Secret fields were therefore serialized into every representation returned to any viewer.

Attack Vector

Exploitation requires network access to the authentik API and valid credentials for an account holding view permission on a configuration that contains secrets. No user interaction is required. The attacker queries the relevant API endpoint and parses the response for secret-bearing fields.

python
# Patch applied to authentik/core/api/utils.py
# The base serializer now strips `secret_fields` from the response
# unless the requesting user holds change_* permission on the instance.
    serializer_field_mapping = BaseModelSerializer.serializer_field_mapping.copy()
    serializer_field_mapping[models.JSONField] = JSONDictField

    def to_representation(self, instance: Model):
        data = super().to_representation(instance)
        # `secret_fields` are visible for users which are allowed to change the object
        secret_fields = getattr(self.Meta, "secret_fields", [])
        request = self.context.get("request")
        if not secret_fields or not request:
            return data
        permission = f"{instance._meta.app_label}.change_{instance._meta.model_name}"
        if request.user.has_perm(permission, instance):
            return data
        for field_name in secret_fields:
            data.pop(field_name, None)
        return data

Source: authentik commit 4b3fad3

Individual serializers were updated to declare their sensitive fields. For example, the Google Workspace provider and Google Chrome connector now mark credentials as a secret_field:

python
# authentik/enterprise/providers/google_workspace/api/providers.py
            "sync_page_timeout",
            "dry_run",
        ]
        secret_fields = ["credentials"]
        extra_kwargs = {}

Source: authentik commit 4b3fad3

Detection Methods for CVE-2026-94611

Indicators of Compromise

  • API GET requests from low-privileged service or operator accounts to endpoints such as /api/v3/sources/, /api/v3/providers/, /api/v3/outposts/, and /api/v3/core/applications/ returning payloads that contain populated secret fields.
  • Unexpected authentication from external systems (Google Workspace, Kubernetes API, SMTP relays, SMS gateways) using credentials associated with authentik integrations.
  • Reuse of authentik-stored OAuth2 client secrets or shared secrets from untrusted IP space.

Detection Strategies

  • Audit authentik access logs for high-volume or scripted enumeration of serializer endpoints by non-administrator accounts.
  • Correlate authentik API read events with downstream credential use on integrated identity sources and provisioning targets.
  • Review Django group and role bindings to identify accounts holding view_* permissions on providers, sources, outposts, and applications without a business justification.

Monitoring Recommendations

  • Enable verbose API access logging on the authentik server and forward it to a centralized log platform for analysis.
  • Alert on API responses containing non-empty values for fields named credentials, client_secret, token, or password when the responding user lacks change permission.
  • Monitor the integrated downstream systems for anomalous authentication using service credentials managed by authentik.

How to Mitigate CVE-2026-94611

Immediate Actions Required

  • Upgrade authentik to 2026.2.7, 2026.5.7, or 2026.8.2 depending on your release train.
  • Rotate every secret stored in affected configurations, including identity source credentials, outbound provisioning tokens, device trust integration keys, Kubernetes outpost tokens, OAuth2/OIDC client secrets, and proxy provider shared secrets.
  • Review and prune view_* permission assignments on provider, source, outpost, and application objects to the minimum set of accounts that require them.

Patch Information

Fixes are available in authentik 2026.2.7, authentik 2026.5.7, and authentik 2026.8.2. The underlying code changes are captured in pull requests #25955, #25960, #25965, and #25970. Full remediation details are in the authentik security advisory GHSA-m9h4-7j9c-55x9.

Workarounds

  • Restrict view permissions on affected object types so that every account with view_* access is also authorized to read the stored secrets, which the advisory confirms is a non-vulnerable configuration.
  • Remove non-essential operator and read-only roles from groups that grant visibility into providers, sources, outposts, and applications until the upgrade is applied.
  • Place the authentik API behind network segmentation that limits reachability to administrative networks while rotation and patching are in progress.
bash
# Example upgrade using the authentik Docker Compose deployment
export AUTHENTIK_TAG=2026.8.2
docker compose pull
docker compose up -d

# Example upgrade using Helm
helm repo update
helm upgrade authentik authentik/authentik \
  --namespace authentik \
  --version 2026.8.2

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

Experience the Most Advanced Cybersecurity Platform

See how the world’s most intelligent, autonomous cybersecurity platform can protect your organization today and into the future.