Skip to main content
CVE Vulnerability Database
Vulnerability Database/CVE-2026-12053

CVE-2026-12053: GitLab EE Information Disclosure Flaw

CVE-2026-12053 is an information disclosure vulnerability in GitLab EE that allows unauthorized access to sensitive committed data. This article covers the technical details, affected versions, impact, and mitigation.

Published:

CVE-2026-12053 Overview

CVE-2026-12053 is an information disclosure vulnerability affecting GitLab Enterprise Edition (EE) versions 19.1 before 19.1.1. The flaw resides in the Duo Workflows feature, where insufficient output filtering allowed unauthorized access to sensitive information previously committed to a project. The weakness is classified under [CWE-532: Insertion of Sensitive Information into Log File].

An unauthenticated network-based attacker can exploit this issue under certain conditions, with no user interaction required. The vulnerability impacts confidentiality across security boundaries, since Duo Workflows surfaces project data outside the intended access scope.

Critical Impact

Sensitive data committed to GitLab projects, including secrets and credentials, may be exposed to unauthorized users through Duo Workflows output.

Affected Products

  • GitLab Enterprise Edition (EE) versions 19.1 through 19.1.0
  • Self-managed GitLab EE instances with Duo Workflows enabled
  • GitLab Duo Workflows component

Discovery Timeline

  • 2026-06-25 - CVE-2026-12053 published to NVD
  • 2026-06-25 - Last updated in NVD database

Technical Details for CVE-2026-12053

Vulnerability Analysis

The vulnerability stems from improper output filtering inside GitLab Duo Workflows, an AI-driven automation feature in GitLab EE. Duo Workflows operates over project data, including repository contents, issues, and historical commits. When generating responses, the workflow engine failed to enforce access boundaries on previously committed content.

Under specific conditions, the workflow output included sensitive material that the requesting user should not have been authorized to view. This includes secrets, tokens, API keys, and any confidential data that had been committed to the project history. The issue maps to [CWE-532], reflecting insertion of sensitive information into accessible output channels.

The attack scope is marked as Changed in the CVSS vector, indicating the vulnerable component leaks data belonging to a different security authority than the one enforcing access. The current Exploit Prediction Scoring System (EPSS) probability stands at 0.328%.

Root Cause

Duo Workflows lacked sufficient output sanitization when retrieving and rendering data sourced from project history. The component returned committed content to the caller without re-evaluating the caller's permissions against the underlying data, breaking the principle of least privilege.

Attack Vector

The attack is network-based and requires no privileges or user interaction. An attacker triggers Duo Workflows operations against a target project, then parses the workflow output to extract sensitive information that was historically committed. Repositories containing credentials, configuration files, or proprietary code are at the highest risk. Refer to the HackerOne Report #3757762 and GitLab Work Item #602194 for detailed technical context.

Detection Methods for CVE-2026-12053

Indicators of Compromise

  • Unexpected Duo Workflows invocations against projects containing sensitive history
  • Workflow outputs containing strings matching secrets patterns such as AKIA, ghp_, xoxb-, or private key headers
  • Abnormal API request volume to Duo Workflows endpoints from unprivileged or anonymous identities
  • Access patterns where users query workflows for projects they have not interacted with previously

Detection Strategies

  • Audit GitLab application logs for Duo Workflows API calls and correlate with the originating user's authorization scope
  • Apply secret-scanning regular expressions against workflow response payloads stored in logs
  • Compare Duo Workflows output references against project membership data to identify unauthorized data exposure

Monitoring Recommendations

  • Enable verbose logging for the Duo Workflows component and forward events to a centralized SIEM
  • Monitor for spikes in workflow execution counts per user and per project
  • Track GitLab audit events tied to project history access and integrate them with identity telemetry

How to Mitigate CVE-2026-12053

Immediate Actions Required

  • Upgrade all GitLab EE instances running 19.1 to version 19.1.1 or later without delay
  • Rotate any credentials, API keys, or tokens that have ever been committed to repositories accessible through Duo Workflows
  • Review Duo Workflows audit logs for evidence of unauthorized data retrieval since the 19.1 release
  • Restrict Duo Workflows access to trusted users until patching is complete

Patch Information

GitLab released the fix in patch release 19.1.1. Administrators should follow the upgrade guidance in the GitLab Patch Release 19.1.1 advisory. The patch corrects the output filtering logic in Duo Workflows so that historical project data is filtered against the caller's effective permissions before being returned.

Workarounds

  • Disable the Duo Workflows feature at the instance or group level until the upgrade to 19.1.1 is applied
  • Remove sensitive content from project history using git filter-repo and invalidate associated secrets
  • Apply stricter group and project visibility settings to limit who can invoke Duo Workflows
bash
# Configuration example - disable Duo Workflows via GitLab Rails console
sudo gitlab-rails console

# Disable Duo Workflows feature flag
Feature.disable(:duo_workflows)

# Verify
Feature.enabled?(:duo_workflows)
# => false

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

Default Legacy - Prefooter | Experience the World’s Most Advanced Cybersecurity Platform

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.