Answer in brief
CVE-2026-75981 records a High severity (CVSS 7.2) vulnerability in TranslatePress – Translate Multilingual sites with AI Translation <= 3.2.5 - Unauthenticated Stored Cross-Site Scripting. The current sources do not mark it as known exploited. The current feed maps cozmoslabs/TranslatePress – Translate Multilingual sites with AI Translation (generic). Check affected ranges and fixed versions before updating.
Analysis pending evidence review
HOL Guard separates source facts from reviewed analysis. See the methodology.
CVSS is 7.2. The current sources do not mark it as known exploited. Treat this as a source-backed prioritization signal, not a statement about your environment.
Analysis status
Analysis pending evidence review
Factual feed record only; HOL analysis is not approved for indexing. Read the methodology.
The current feed maps cozmoslabs/TranslatePress – Translate Multilingual sites with AI Translation (generic). Check affected ranges and fixed versions before updating.
| Package | Affected range | Fixed version |
|---|---|---|
| cozmoslabs/TranslatePress – Translate Multilingual sites with AI Translationgeneric | 0 | Not reported |
Published upstream
Aug 19, 2026
Evidence: source:cvelist:source_dates:source-dates:recordSource modified
Aug 20, 2026
Evidence: source:cvelist:source_dates:source-dates:recordFirst seen by HOL
Aug 19, 2026
The TranslatePress – Translate Multilingual sites with AI Translation plugin for WordPress is vulnerable to unauthenticated Stored Cross-Site Scripting in versions up to and including 3.2.5. The special gettext markers '#!trpst#' and '#!trpen#' are unconditionally rewritten to '<' and '>' by translate_page() in includes/class-translation-render.php (lines 538-539). Because those markers are plain text with no HTML-special characters, an unauthenticated attacker can embed them in a comment; the markers survive wp_kses, and when the post is viewed in a secondary language the substitution turns the attacker's '#!trpst#img ... #!trpen#' into a real <img> tag. remove_tags_from_output() only strips <script>/<style>, so an <img onerror=...> executes in the visitor's browser.
Quoted source text, attributed separately from HOL analysis.