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-2026-84911 Overview
CVE-2026-84911 is a stored cross-site scripting (XSS) vulnerability [CWE-79] in the Drupal AI (Artificial Intelligence) contributed module. The flaw stems from improper neutralization of user-supplied input during web page generation. An authenticated attacker with high privileges can inject persistent script payloads that execute in the browsers of users who interact with affected pages.
The issue affects AI (Artificial Intelligence) module versions 0.0.0 through 1.3.13, and 1.4.0 through 1.4.8. Drupal has published advisory SA-2026-119 describing the vulnerability and remediation.
Critical Impact
Stored script payloads can hijack authenticated sessions, trigger unauthorized actions, or alter rendered content for users who load the affected pages.
Affected Products
- Drupal AI (Artificial Intelligence) module versions 0.0.0 to 1.3.13
- Drupal AI (Artificial Intelligence) module versions 1.4.0 to 1.4.8
- Drupal sites running the AI contributed module with privileged content authoring enabled
Discovery Timeline
- 2026-10-22 - CVE-2026-84911 published to NVD
- 2026-10-22 - Last updated in NVD database
Technical Details for CVE-2026-84911
Vulnerability Analysis
The Drupal AI module fails to sanitize or encode specific user-controlled input before rendering it in generated web pages. When a privileged user submits crafted content containing HTML or JavaScript, the module stores the payload and later returns it to viewers without proper neutralization. Browsers then parse the attacker-supplied markup as executable script within the site's origin.
Exploitation requires an authenticated account with high privileges and user interaction from a victim who loads the affected view. The attack reaches across the network and executes with the privileges of the viewing user's browser session. Impact is limited to confidentiality and integrity of the affected session, with no direct availability impact.
Root Cause
The root cause is missing output encoding in rendering paths that handle AI-related content fields. The module trusts input that passed server-side storage validation, bypassing Drupal's standard sanitization filters such as Xss::filter() or the Twig auto-escape layer. Any field feeding into these render paths becomes a stored injection sink.
Attack Vector
An attacker with elevated module permissions submits a payload containing script content through an AI module field. The payload persists in the Drupal database. When another user, including administrators, visits a page rendering that field, the browser executes the attacker's script in the context of the Drupal site, enabling session theft, forced actions via CSRF-style requests, or UI manipulation.
No verified exploit code is publicly available. See the Drupal Security Advisory SA-2026-119 for vendor-provided technical details.
Detection Methods for CVE-2026-84911
Indicators of Compromise
- Unexpected <script>, <iframe>, or event-handler attributes (e.g., onerror, onload) stored in AI module content fields or configuration entities
- Outbound HTTP requests from authenticated administrator browsers to attacker-controlled domains shortly after rendering AI module pages
- New or modified Drupal user accounts or role assignments created shortly after an admin viewed AI module content
Detection Strategies
- Query the Drupal database for AI module fields containing HTML tags, JavaScript URI schemes (javascript:), or encoded script fragments
- Review web server access logs for POST requests to AI module endpoints from accounts with high privileges, correlated with unusual follow-on GET traffic
- Inspect browser Content Security Policy (CSP) violation reports for inline script executions originating from AI module routes
Monitoring Recommendations
- Enable Drupal's watchdog logging for the AI module and forward events to a centralized log platform for retention and search
- Alert on administrative role changes, permission grants, or node edits performed immediately after an admin viewed AI module content
- Monitor outbound requests from staff workstations to newly registered or low-reputation domains following AI module page loads
How to Mitigate CVE-2026-84911
Immediate Actions Required
- Upgrade the Drupal AI (Artificial Intelligence) module to a fixed release as specified in Drupal Security Advisory SA-2026-119
- Audit existing AI module content and configuration for stored HTML or script payloads and remove any injected markup
- Review which accounts hold the high-privilege permissions required to exploit this issue and remove unnecessary grants
Patch Information
Upgrade the AI (Artificial Intelligence) module beyond versions 1.3.13 and 1.4.8 per the fixed releases listed in SA-2026-119. Clear Drupal caches after upgrading and re-save affected content so sanitized rendering takes effect. Refer to the Drupal Security Advisory SA-2026-119 for the authoritative version matrix.
Workarounds
- Restrict the elevated AI module permissions to a minimal, trusted set of users until the upgrade is applied
- Deploy a strict Content Security Policy that disallows inline scripts (script-src 'self') to reduce exploitability of stored XSS payloads
- Place the Drupal site behind a web application firewall rule set that blocks obvious script and event-handler patterns submitted to AI module endpoints
# Configuration example: upgrade the AI module via Composer and clear caches
composer update drupal/ai --with-dependencies
drush updatedb -y
drush cache:rebuild
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

