Skip to main content
Vulnerability Database/CVE-2026-114729

CVE-2026-114729: Roundcube Webmail Content Blocking Bypass

CVE-2026-114729 is a remote content blocking bypass in Roundcube Webmail that allows attackers to load remote content without user consent, exposing client IP addresses. This article covers technical details, affected versions, security impact, and mitigation strategies.

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-2026-114729 Overview

CVE-2026-114729 is a remote content blocker bypass in Roundcube Webmail affecting versions before 1.6.20 and 1.7.x before 1.7.5. The HTML sanitizer rcube_washtml validated the SVG SMIL animation attributes to, from, and values for URIs, but failed to inspect the by attribute. Attackers can craft HTML email messages containing SVG SMIL markup that uses the by attribute to reference an external URI. When a recipient opens the message, the webmail client loads the remote content without consent, disclosing the client IP address and confirming message viewing. The weakness is categorized as [CWE-184] (incomplete list of disallowed inputs).

Critical Impact

Attackers can confirm email opening and harvest recipient IP addresses by sending a crafted HTML message, enabling tracking, reconnaissance, and phishing effectiveness measurement.

Affected Products

  • Roundcube Webmail versions prior to 1.6.20
  • Roundcube Webmail 1.7.x versions prior to 1.7.5
  • Any downstream distribution packaging the affected versions (for example, Debian — see bug report #1151279)

Discovery Timeline

  • 2026-10-19 - Roundcube publishes security updates 1.6.20 and 1.7.5
  • 2026-10-22 - CVE-2026-114729 published to NVD
  • 2026-10-22 - Last updated in NVD database

Technical Details for CVE-2026-114729

Vulnerability Analysis

Roundcube's rcube_washtml HTML sanitizer enforces the remote-content blocker by inspecting attributes that may contain URIs and stripping or neutralizing external references. For SVG SMIL animation elements, the sanitizer specifically checked the to, from, and values attributes because they can carry arbitrary content, including URI references used by set and animate elements.

The by attribute was omitted from this allowlist check. SVG SMIL permits by on animation elements to express an animation delta, and when combined with attributeName="href" on an element such as <image> or <feImage>, it can resolve to a remote URL. The sanitizer passed the attribute through unmodified, so the browser fetched the referenced resource when the message was rendered.

Root Cause

The root cause is an incomplete denylist in attribute inspection logic. The URI-bearing-attribute check in program/lib/Roundcube/rcube_washtml.php enumerated three attribute names but did not include by, which shares the same URI-permitting semantics in SVG SMIL animations.

Attack Vector

Exploitation requires only that a victim open a crafted HTML email in a vulnerable Roundcube session. No authentication to the attacker's infrastructure, no additional user interaction, and no scripting are required. The attacker's web server records the fetch, revealing the IP address, user agent, and timestamp of the read event.

php
                $out   = null;

                // in SVG to/from attribs may contain anything, including URIs
-                if ($key == 'to' || $key == 'from' || $key == 'values') {
+                if ($key == 'to' || $key == 'from' || $key == 'values' || $key == 'by') {
                    $key = strtolower((string) $node->getAttribute('attributeName'));
                    $key = trim(preg_replace('/^.*:/', '', $key));
                    if ($key && !isset($this->_html_attribs[$key])) {

Source: Roundcube Commit 2b6ff79 and Roundcube Commit 96d993b. The patch extends the URI-bearing attribute check to also cover by, routing it through the same attributeName-based sanitization used for the other SMIL animation attributes.

Detect Identity Attacks the Moment Authentication Fails

Singularity™ Identity surfaces credential misuse, lateral movement, and privilege escalation in real time, so you can contain the breach before it spreads.

Stop Identity Attacks

Detection Methods for CVE-2026-114729

Indicators of Compromise

  • Inbound HTML email messages containing SVG elements with <animate>, <set>, or <animateTransform> tags that use a by attribute referencing an external URL.
  • Outbound HTTP(S) requests from webmail client IP addresses to unknown or newly registered domains occurring immediately after a message is opened.
  • Web server logs showing GET requests with referrers pointing to the Roundcube message rendering endpoint.

Detection Strategies

  • Deploy mail gateway content inspection that parses MIME HTML parts and flags SVG SMIL animation attributes carrying URI schemes (http:, https:, //).
  • Correlate webmail session activity with outbound DNS and HTTP connections to detect unexpected external fetches triggered by message rendering.
  • Review Roundcube and reverse proxy access logs for message-view events immediately followed by distinct external connections from the same client.

Monitoring Recommendations

  • Monitor the installed Roundcube version in your deployment and alert on versions below 1.6.20 or 1.7.5.
  • Instrument egress monitoring from webmail servers to identify beaconing patterns consistent with tracking pixels delivered via SVG.
  • Track user-reported phishing with HTML attachments and extract SVG payloads for retrospective analysis.

How to Mitigate CVE-2026-114729

Immediate Actions Required

  • Upgrade Roundcube Webmail to 1.6.20 or 1.7.5 as published in the Roundcube Security Updates Announcement.
  • Apply distribution patches where available, such as the fix referenced in Debian Bug Report #1151279.
  • Audit webmail logs for message-open events preceding external callouts and notify affected users if tracking is suspected.

Patch Information

The fix is included in Roundcube Release 1.6.20 and Roundcube Release 1.7.5. The code change adds the by attribute to the URI-bearing attribute allowlist in rcube_washtml, forcing it through the same sanitization applied to to, from, and values.

Workarounds

  • Keep the Roundcube default of blocking remote content enabled and instruct users not to click "Load remote images" for untrusted senders — note this does not fully mitigate the SMIL fetch triggered during initial render.
  • Deploy an upstream HTML/SVG sanitizer at the mail gateway that strips SVG SMIL animation elements from inbound messages until patches are applied.
  • Restrict outbound HTTP egress from the webmail server to a documented allowlist to limit the value of tracking callbacks.
bash
# Verify installed Roundcube version and upgrade on Debian-based systems
dpkg -l | grep roundcube
apt-get update && apt-get install --only-upgrade roundcube roundcube-core

# For source installs, download and deploy the fixed release
wget https://github.com/roundcube/roundcubemail/releases/download/1.6.20/roundcubemail-1.6.20-complete.tar.gz
tar xzf roundcubemail-1.6.20-complete.tar.gz
# Follow the UPGRADING instructions in the extracted directory

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.