CVE-2026-18978

HighCVSS 7.2

Published: 28/08/2026(UTC)Updated: 28/08/2026(UTC)Translated: 28/08/2026(UTC)NVD NIST

Exploitation Probability (EPSS)

Low risk

0.27%

19th percentile - higher than 19% of all known CVEs

Summary

The LiteSpeed Cache plugin for WordPress up to version 7.8.1 is vulnerable to Stored Cross-Site Scripting via Comment Content. Insufficient input sanitization and output escaping allow unauthenticated attackers to inject arbitrary web scripts that execute when a user accesses an injected page. A payload crafted from decimal numeric character references bypasses WordPress's wp_kses sanitization.

Risk Assessment

Attackers can inject malicious scripts, potentially leading to data theft, session hijacking, or malware distribution. The risk includes compromise of user security and site integrity.

Recommendation

Update the LiteSpeed Cache plugin to the latest patched version. Additionally, consider disabling comments or tightening moderation policies.

Other vulnerabilities in LiteSpeed Cache

CVE-2026-3129Medium
The LiteSpeed Cache plugin for WordPress is vulnerable to Stored Cross-Site Scripting via crafted `<img>` tag attributes in all versions up to, and including, 7.7. This is due to a flawed regular expression used to strip `width` and `height` attributes from images when the "Lazy Load Images" and "Add Missing Sizes" features are enabled.

Original NVD description (English source)

The LiteSpeed Cache plugin for WordPress is vulnerable to Stored Cross-Site Scripting via Comment Content in all versions up to, and including, 7.8.1 due to insufficient input sanitization and output escaping. This makes it possible for unauthenticated attackers to inject arbitrary web scripts in pages that will execute whenever a user accesses an injected page. A comment payload crafted exclusively from decimal numeric character references (e.g. ", <, >) placed inside an allowed element such as <code> bypasses WordPress's wp_kses sanitization, as kses does not treat a data-settings="..." substring within text content as an HTML attribute, allowing the malicious payload to reach the vulnerable function. For this to be exploitable, the site must allow users with previously approved comments to write new comments, and the require_name_email setting must be disabled.