Description

The AI Builder WordPress plugin before 2.7.8 does not sanitise custom JavaScript saved against a post before echoing it inside a script tag on the front end, allowing users with contributor level access and above to store arbitrary JavaScript that will execute in the browser of anyone who views the post, including the editor or administrator who reviews it.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "6.8",
  "pubDate": "2026-09-11T07:16:47.247Z",
  "pubdate": "2026-09-11T07:16:47.247Z",
  "executiveSummary": "The AI Builder WordPress plugin, in versions prior to 2.7.8, contains a Stored Cross-Site Scripting (XSS) vulnerability due to improper input sanitization. This flaw allows authenticated users with at least 'contributor' level permissions to inject and store malicious JavaScript payloads within post content. When these posts are rendered on the frontend, the injected script executes within the context of the victim's browser session. The primary risk involves the compromise of administrative or editorial sessions, leading to unauthorized actions, data exfiltration, or the deployment of further malicious content. The vulnerability is characterized by a high impact on integrity and confidentiality, as it enables attackers to manipulate the content rendered to privileged users. Successful exploitation requires an attacker to possess valid contributor-level credentials, bypassing security boundaries intended to restrict script injection. As the malicious payload is stored permanently, it persists until explicitly removed, posing a continuous threat to any user viewing the affected post.",
  "technicalDetails": "The root cause of this vulnerability is the failure of the AI Builder plugin to implement sufficient input sanitization or output encoding for custom JavaScript fields associated with post data. In versions 2.7.7 and below, the plugin accepts custom JavaScript input from users and stores it directly in the WordPress database without validating the content against a whitelist or sanitizing it for dangerous HTML tags and script execution vectors.\nThe attack flow begins when an authenticated user with contributor access—or higher—navigates to the AI Builder interface provided by the plugin. The attacker embeds a crafted JavaScript payload into the custom JavaScript configuration field for a specific post. Because the plugin lacks necessary server-side filtering, the malicious payload is saved to the post metadata or designated database table.\nThe vulnerability manifests during the frontend rendering phase. When a visitor, including high-privileged users such as site administrators or editors, accesses the published post, the plugin retrieves the stored, unsanitized JavaScript from the database. It then reflects this content directly into an inline <script> tag within the HTML document sent to the victim's browser. As the browser parses the rendered page, the script executes immediately within the security context of the origin.\nThis execution leads to a classic Stored XSS scenario. The payload can perform various actions, such as capturing session cookies, hijacking administrative sessions, performing unauthorized requests on behalf of the victim (CSRF), or redirecting users to malicious external sites. Since the payload is triggered whenever a victim views the page, an attacker can target site administrators specifically by prompting them to review a post containing the payload, thereby escalating their own privileges or gaining full control over the WordPress instance. This flaw exists due to a fundamental breakdown in the trust boundary between user-supplied content and the application's output rendering mechanism, failing to adhere to the principle of never trusting client-side input."
}