CVE-2026-25517: Wagtail CMS Information Disclosure Flaw

CVE-2026-25517 Overview

CVE-2026-25517 is a Missing Authorization vulnerability (CWE-862) affecting Wagtail, an open source content management system built on Django. Due to a missing permission check on the preview endpoints, authenticated users with access to the Wagtail admin interface can craft form submissions to obtain preview renderings of pages, snippets, or site settings without proper authorization.

This authorization bypass allows attackers to render arbitrary data through Wagtail's preview functionality, potentially exposing database contents that would normally require edit-level access to view.

Critical Impact

Authenticated admin users can bypass authorization controls to preview arbitrary model data, potentially leaking sensitive database contents through template rendering without requiring edit permissions.

Affected Products

• Wagtail versions prior to 6.3.6
• Wagtail versions 7.0.x prior to 7.0.4
• Wagtail versions 7.1.x prior to 7.1.3
• Wagtail versions 7.2.x prior to 7.2.2
• Wagtail versions prior to 7.3

Discovery Timeline

• 2026-02-04 - CVE-2026-25517 published to NVD
• 2026-02-05 - Last updated in NVD database

Technical Details for CVE-2026-25517

Vulnerability Analysis

This vulnerability stems from inadequate authorization enforcement in Wagtail's preview endpoint functionality. Wagtail provides preview capabilities for pages, snippets, and site settings, allowing content editors to see how changes will appear before publishing. However, the affected versions fail to verify that the requesting user has appropriate permissions to access or edit the specific model being previewed.

The authorization gap allows any authenticated user with basic Wagtail admin access to construct malicious form submissions targeting preview endpoints. By manipulating the request parameters, an attacker can inject arbitrary data into the preview rendering pipeline. While this does not directly expose the existing stored data of the target object, the template rendering process may inadvertently reveal other database contents through template context variables, related object lookups, or dynamic content inclusion.

The attack requires knowledge of the target model's field structure to craft a valid preview request. This information could be obtained through reconnaissance, documentation review, or prior legitimate access to similar models.

Root Cause

The root cause is a Missing Authorization vulnerability (CWE-862) where the preview endpoints fail to implement proper permission checks before processing preview requests. The application does not verify that the authenticated user has the necessary permissions (such as add, edit, or view rights) for the specific model type being previewed, relying solely on basic admin authentication.

Attack Vector

The attack is network-based and requires an authenticated session with Wagtail admin access. An attacker must:

1. Authenticate to the Wagtail admin interface with any valid credentials
2. Identify a target model type that supports preview functionality
3. Determine the field structure of the target model
4. Craft a POST request to the preview endpoint with attacker-controlled data
5. Analyze the rendered preview response for potentially sensitive information

The vulnerability can be exploited by crafting HTTP POST requests to preview endpoints with manipulated form data. Attackers can include arbitrary field values that, when rendered through the model's template, may trigger lookups or display logic that exposes protected data. The preview functionality renders the submitted data as if it were a legitimate object, potentially including related objects, computed properties, or other database content in the output.

For detailed technical information about the vulnerability mechanism, refer to the GitHub Security Advisory GHSA-4qvv-g3vr-m348.

Detection Methods for CVE-2026-25517

Indicators of Compromise

• Unusual or high-volume POST requests to Wagtail preview endpoints from users without edit permissions
• Preview requests containing field values that reference objects the user should not have access to
• Authentication logs showing admin users accessing preview functionality for models outside their normal scope

Detection Strategies

• Monitor Wagtail admin access logs for preview endpoint requests (/admin/pages/*/preview/, /admin/snippets/*/preview/) from users without corresponding edit permissions
• Implement application-layer logging to capture the model types and user context for all preview operations
• Deploy web application firewall rules to flag preview requests with suspicious or unexpected field patterns

Monitoring Recommendations

• Enable detailed Django request logging for the Wagtail admin namespace
• Configure alerting for preview endpoint access by users lacking model-specific permissions
• Review audit logs periodically for patterns of preview abuse across different model types

How to Mitigate CVE-2026-25517

Immediate Actions Required

• Upgrade Wagtail to patched versions: 6.3.6, 7.0.4, 7.1.3, 7.2.2, or 7.3 or later
• Audit admin user accounts and restrict access to only those who require it
• Review recent access logs for any suspicious preview endpoint activity

Patch Information

Wagtail has released patched versions that address this vulnerability by implementing proper permission checks on preview endpoints:

• Version 6.3.6 for the 6.3.x branch
• Version 7.0.4 for the 7.0.x branch
• Version 7.1.3 for the 7.1.x branch
• Version 7.2.2 for the 7.2.x branch
• Version 7.3 and later

The fix ensures that users must have appropriate permissions (add or edit) for the target model before preview requests are processed. Multiple commits implement this security improvement:

• Wagtail Commit Fix
• Wagtail Commit Patch
• Wagtail Commit Change

Workarounds

• Restrict Wagtail admin access to only trusted users who require content management capabilities
• Disable preview functionality for sensitive models by overriding the preview_modes attribute
• Implement network-level access controls to limit admin interface access to trusted IP ranges

# Upgrade Wagtail to a patched version
pip install --upgrade wagtail>=7.3

# Or for specific branch versions:
pip install wagtail==6.3.6
pip install wagtail==7.0.4
pip install wagtail==7.1.3
pip install wagtail==7.2.2