CVE-2026-118551: Drupal AI module information disclosure in speech-to-text audio

CVE-2026-118551 · Severity: info · Published 2026-10-30

Drupal's AI integration module writes audio files from speech-to-text requests to the temporary directory without deleting them, allowing attackers with read access to that directory to retrieve transcription audio from other users. This creates a data exposure risk for sensitive audio content, though exploitation requires the attacker to already have server or directory access through shared hosting or other exposed file mechanisms.

Technologies: Drupal AI, Packagist:Https://Packages.Drupal.Org/8 Drupal/Ai. Vendors: Drupal, Packagist:Https://Packages.Drupal.Org/8.

Quick answers

Is CVE-2026-118551 exploited in the wild?
No. Junglewise Threat Intelligence has no report of CVE-2026-118551 being exploited in the wild.
How severe is CVE-2026-118551?
Informational severity, published 30 October 2026.
Which products and versions are affected?
Drupal AI all versions before 1.4.11; 1.5.0 before 1.5.2; Packagist:Https://Packages.Drupal.Org/8 Drupal/Ai ECOSYSTEM:introduced=0,fixed=1.4.11|ECOSYSTEM:introduced=1.5.0,fixed=1.5.2.
Is there a fix for CVE-2026-118551?
Yes. Packagist:Https://Packages.Drupal.Org/8 Drupal/Ai: upgrade to 1.4.11 or later. Until then: Delete any leftover speech_to_text.mp3 files from the site's temporary directory to remove already-exposed audio. Junglewise finds every place you run it, drafts the upgrade and applies it once you approve.
How is CVE-2026-118551 attacked?
It is attacked with local access or from the adjacent network, in a software library.
Am I affected by CVE-2026-118551?
Check whether you run Drupal AI or Packagist:Https://Packages.Drupal.Org/8 Drupal/Ai at an affected version. Junglewise does that check for you around the clock: connect your code, cloud and domains, and it matches new vulnerabilities like this one against what you actually run and tells you when one reaches you.

How to fix

  • Packagist:Https://Packages.Drupal.Org/8 Drupal/Ai: upgrade to 1.4.11 or later

Delete any leftover speech_to_text.mp3 files from the site's temporary directory to remove already-exposed audio.

Technical details

The vulnerability is an information disclosure flaw where the AI module persists audio files in a shared temporary directory without cleanup. An attacker with local or adjacent access to the temporary directory (via shared hosting, exposed file endpoints, or privilege escalation) can recover speech-to-text audio intended for other users. The fix moves audio handling to memory-only, eliminating filesystem persistence.

Affected products

  • Drupal AI all versions before 1.4.11; 1.5.0 before 1.5.2

Identifiers

  • CVE-2026-118551
  • DRUPAL-CONTRIB-2026-271

Timeline

  • 2026-10-28: disclosed

References

This page as Markdown. Free to reuse under CC BY 4.0 with a link back.