CVE-2026-27013

fabric.js applies escapeXml() to text content during SVG export (src/shapes/Text/TextSVGExportMixin.ts:186) but fails to apply it to other user-controlled string values that are interpolated into SVG attribute markup. When attacker-controlled JSON is loaded via loadFromJSON() and later exported via toSVG(), the unescaped values break out of XML attributes and inject arbitrary SVG elements including event handlers.

Deserialization Path (no sanitization)

loadFromJSON() (src/canvas/StaticCanvas.ts:1229) calls enlivenObjects() which calls _fromObject() (src/shapes/Object/Object.ts:1902). _fromObject passes all deserialized properties to the shape constructor via new this(enlivedObjectOptions). The constructor ultimately calls _setOptions() (src/CommonMethods.ts:9) which iterates over every property and assigns it to the object via this.set(prop, options[prop]). There is no allowlist or sanitization - any property in the JSON, including id, is set verbatim on the fabric object.

Finding 1: XSS via id Property Injection

The id property from deserialized JSON is interpolated directly into SVG attribute strings without escaping.

Vulnerable code (src/shapes/Object/FabricObjectSVGExportMixin.ts, line 89, getSvgCommons()):

getSvgCommons(
  this: FabricObjectSVGExportMixin & FabricObject & { id?: string },
) {
  return [
    this.id ? `id="${this.id}" ` : '',  // <-- unescaped, user-controlled
    this.clipPath
      ? `clip-path="url(#${...})" `
      : '',
  ].join('');
}

This method is called in _createBaseSVGMarkup() (same file, line 178) which wraps every object's SVG output in a <g> element. Every fabric object type (Rect, Circle, Path, Text, Image, Group, etc.) inherits this mixin, so the id injection vector applies to all object types.

Contrast with text content, which IS escaped:

// src/shapes/Text/TextSVGExportMixin.ts:186
return `<tspan ...>${escapeXml(char)}</tspan>`;

The inconsistency shows that the intention was to prevent injection but was missed w attribute contexts.

Finding 2: XSS via Image src / xlink:href Injection

Image source URLs are interpolated raw into xlink:href in _toSVG().

Vulnerable code (src/shapes/Image.ts, line 404, _toSVG()):

imageMarkup.push(
  '\t<image ',
  'COMMON_PARTS',
  `xlink:href="${this.getSvgSrc(true)}" x="${x - this.cropX}" y="${
    y - this.cropY
  }" ...`  // <-- unescaped
);

getSvgSrc() returns the image src property which is set from JSON during deserialization. An attacker can inject a src value that breaks out of the xlink:href attribute.

Finding 3: XSS via Pattern sourceToString()

Vulnerable code (src/Pattern/Pattern.ts, line 181, toSVG()):

`<image x="0" y="0" ... xlink:href="${this.sourceToString()}"></image>`
// <-- unescaped, returns this.source.src for image sources

Additionally, Pattern's constructor (line 92–94) runs this.id = uid() before Object.assign(this, options), meaning a user-supplied id in the pattern JSON overwrites the auto-generated uid. The pattern id is then interpolated unescaped on line 180:

`<pattern id="SVGID_${id}" x="${patternOffsetX}" ...>`

Finding 4: Gradient id Partial Injection (lower Severity)

Vulnerable code (src/gradient/Gradient.ts, line 212, toSVG()):

`id="SVGID_${this.id}"`  // <-- unescaped

Gradient's constructor (line 125) computes id: id ? ${id}_${uid()} : uid(). If a user-supplied id is present in the gradient JSON, it is prepended to the auto-generated uid. The user-controlled portion is interpolated unescaped into the SVG. This is exploitable but the payload is constrained by the _<uid> suffix appended after it.

Confirmed Affected Files

Any application that:

1. Accepts user-supplied JSON (via loadFromJSON(), collaborative sharing, import features, CMS plugins), AND
2. Renders the toSVG() output in a browser context (SVG preview, export download rendered in-page, email template, embed)

...is vulnerable to stored XSS. An attacker can execute arbitrary JavaScript in the victim's browser session.

Real-world attack scenarios:

• Collaborative design tools (Canva-like apps) where users share canvas state as JSON
• CMS or e-commerce platforms with fabric.js-based editors that store/render designs
• Any export-to-SVG workflow where the SVG is later displayed in a browser

Untrusted input is rendered as active markup in a victim's browser, which can run script in their session. Typical impact: session or credential theft, and actions taken as the user.

CVE-2026-27013 has a CVSS score of 7.6 (High). The vector is network-reachable, no privileges required, and user interaction required. A CVSS score reflects the worst-case severity of the vulnerability, not your specific exposure. Whether this affects your application depends on whether the vulnerable code is present and reachable in your environment. A fixed version is available (7.2.0); upgrading removes the vulnerable code path.

Kodem intelligence

Update to fabric.js 7.2.0 or newer version.

1. What is CVE-2026-27013? CVE-2026-27013 is a high-severity cross-site scripting (XSS) vulnerability in fabric (npm), affecting versions < 7.2.0. It is fixed in 7.2.0. Untrusted input is rendered as active markup in a victim's browser, which can run script in their session.
2. How severe is CVE-2026-27013? CVE-2026-27013 has a CVSS score of 7.6 (High). This score reflects the worst-case severity of the vulnerability, not your specific exposure. Whether it represents real risk in your environment depends on whether the vulnerable code is present and reachable.
3. Which versions of fabric are affected by CVE-2026-27013? fabric (npm) versions < 7.2.0 is affected.
4. Is there a fix for CVE-2026-27013? Yes. CVE-2026-27013 is fixed in 7.2.0. Upgrade to this version or later.
5. Is CVE-2026-27013 exploitable, and should I be worried? Whether CVE-2026-27013 is exploitable in your environment depends on whether the vulnerable code is present and reachable. A CVSS score is a worst-case rating; it does not account for your specific deployment, configuration, or usage patterns. Kodem, an Intelligent Application Security platform, uses runtime intelligence to show which vulnerabilities actually execute in production, so you can focus on the ones that represent real risk. Get a demo
6. What actually determines whether CVE-2026-27013 is exploitable, and how bad it is? Exploitability and impact are not fixed properties of a CVE. They depend on runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A high CVSS score on a dependency that never runs is not the same as real risk. Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter.
7. How do I fix CVE-2026-27013? Upgrade fabric to 7.2.0 or later.