CVE-2026-1400: 
WordPress vulnerability analysis and mitigation

CVE-2026-1400 is an arbitrary file upload vulnerability in the AI Engine – The Chatbot and AI Framework for WordPress plugin, affecting all versions up to and including 3.3.2. The flaw stems from missing file type validation in the rest_helpers_update_media_metadata function, enabling authenticated attackers with Editor-level access or higher to upload arbitrary files — including PHP web shells — to the server. It was published on January 28, 2026, and assigned a CVSS v3.1 base score of 7.2 (High) (Wordfence, Red Hat CVE).

The vulnerability is classified as CWE-434 (Unrestricted Upload of File with Dangerous Type). The attack exploits a two-step process: an attacker first uploads a benign image file through normal WordPress media upload mechanisms, then calls the update_media_metadata REST API endpoint to rename the file to a .php extension, effectively placing an executable PHP file in the uploads directory. Because the rest_helpers_update_media_metadata function performs no validation on the resulting file type after renaming, the server will execute the file as PHP upon HTTP request. Exploitation requires network access and at minimum Editor-level WordPress credentials (Wordfence, WordPress Trac).

Successful exploitation allows an authenticated attacker to achieve remote code execution (RCE) on the underlying web server, with the privileges of the web server process. This can lead to full compromise of the WordPress site, including theft of sensitive data (database credentials, user PII), defacement, installation of persistent backdoors, and potential lateral movement to other services hosted on the same server. Confidentiality, integrity, and availability are all rated as High impact (Wordfence).

No public exploit code or active in-the-wild exploitation has been reported as of the available data. The EPSS score is approximately 0.076%, indicating a low current probability of exploitation in the wild. The vulnerability is not listed in the CISA Known Exploited Vulnerabilities (KEV) catalog. Exploitation requires authenticated access at the Editor level or above, which limits the attack surface compared to unauthenticated vulnerabilities (Wordfence, Red Hat CVE).

1. Reconnaissance: Identify WordPress sites running the AI Engine plugin (version ≤ 3.3.2) by checking plugin metadata via /wp-content/plugins/ai-engine/readme.txt or using tools like WPScan.
2. Obtain Editor credentials: Acquire valid WordPress credentials with Editor-level access or higher through phishing, credential stuffing, or other means.
3. Upload a benign image: Use the WordPress media upload interface or REST API to upload a legitimate image file (e.g., shell.jpg) containing embedded PHP code (e.g., <?php system($_GET['cmd']); ?>).
4. Rename via vulnerable endpoint: Send an authenticated POST request to the update_media_metadata REST API endpoint, specifying the uploaded file's ID and setting the filename to shell.php.
5. Trigger RCE: Access the renamed file at its URL in the uploads directory (e.g., https://target.com/wp-content/uploads/YYYY/MM/shell.php?cmd=id) to execute arbitrary OS commands (Wordfence, WordPress Trac).

• Network: HTTP POST requests to /wp-json/ endpoints associated with update_media_metadata or rest_helpers_update_media_metadata from authenticated sessions; outbound connections from the web server process to unknown external IPs.
• File System: Presence of .php files in the WordPress uploads directory (/wp-content/uploads/) that were not created by normal WordPress operations; files with embedded PHP code disguised as image content.
• Logs: WordPress access logs showing REST API calls to media metadata update endpoints followed shortly by GET requests to .php files in the uploads directory; PHP error logs referencing unexpected script execution in the uploads path.
• Process: Unusual child processes spawned by the web server (e.g., Apache/Nginx/PHP-FPM) such as bash, curl, wget, or python with suspicious arguments.

Update the AI Engine plugin to version 3.3.3 or later, which includes a patch that adds proper file type validation in the rest_helpers_update_media_metadata function (WordPress Trac). As a temporary workaround, restrict Editor-level and above accounts to trusted users only, and consider using a Web Application Firewall (WAF) rule to block REST API requests that attempt to rename media files to executable extensions. Additionally, configure the web server to deny execution of PHP files within the wp-content/uploads/ directory as a defense-in-depth measure (Wordfence).

Wordfence, which discovered and reported the vulnerability, published it in their weekly WordPress vulnerability report for the period of January 26–February 1, 2026 (Wordfence Blog). The vulnerability received routine coverage from security aggregators including RedPacketSecurity on Mastodon and Twitter/X, as well as Spanish national cybersecurity agencies INCIBE-CERT and CCN-CERT. Community reaction has been measured, consistent with the High (but not Critical) severity and the requirement for authenticated access.