Loading vulnerability…
CVE-2023-51409 is an unrestricted upload of file with dangerous type vulnerability in the Jordy Meow AI Engine: ChatGPT Chatbot WordPress plugin. The issue affects versions through 1.9.98. Based on the available information, the plugin does not sufficiently restrict uploaded file types, allowing an attacker to upload dangerous files. If the application or underlying server subsequently stores those files in a web-accessible location or otherwise processes them insecurely, the flaw could be leveraged to deliver malicious content and potentially achieve further compromise.
Mallory correlates every CVE against your assets, your vendors, and active adversary campaigns. Know which vulnerabilities matter for you, not just which ones are loud.
What it means. What to do now. Patch path, mitigations, and the assume-compromise checklist.
What an attacker gets, and what they’ve been doing with it.
If you can’t patch tonight, do this now.
Patch, then assume compromise.
2 valid exploits after Mallory filtered fakes, detection scripts, and README-only repos.
Repository contains an operational Python exploit (CVE-2023-51409.py) targeting an unauthenticated arbitrary file upload in the WordPress plugin “AI Engine: ChatGPT Chatbot”. The exploit posts a multipart upload to the REST endpoint /wp-json/mwai-ui/v1/files/upload, uploading a PHP webshell (cmd.php) that executes system($_GET['cmd']). It then assumes WordPress’ default uploads path (/wp-content/uploads/YYYY/MM/) to locate the dropped file, validates accessibility, performs simple OS detection by reading the webshell’s output, and provides an interactive loop to run arbitrary commands via HTTP GET requests. Repo structure: (1) CVE-2023-51409.py (main exploit entry point) and README.md (usage/description). (2) A large plugin/1.9.98/ directory containing a snapshot of the AI Engine plugin source (PHP/JS/CSS) including the vulnerable REST route implementation in classes/modules/files.php where /mwai-ui/v1/files/upload is registered with permission_callback '__return_true' and uses move_uploaded_file/media_handle_upload to store uploaded files. Numerous .svn/pristine artifacts mirror the same plugin files. Key capability: pre-auth network RCE by uploading executable PHP into a web-accessible uploads directory. No C2 infrastructure is embedded; interaction is direct to the target WordPress site over HTTP(S).
This repository contains a Python exploit script (CVE-2023-51409.py) targeting the AI Engine: ChatGPT Chatbot WordPress plugin (versions <= 1.9.98). The exploit first checks the plugin version by fetching /wp-content/plugins/ai-engine/readme.txt from the target site. If the version is vulnerable, it proceeds to upload an arbitrary PHP file (default: a webshell that runs 'id') to the /wp-json/mwai-ui/v1/files/upload REST endpoint. Upon successful upload, the PHP file is accessible in the /wp-content/uploads/ directory, allowing remote code execution. The exploit is unauthenticated and leverages a network attack vector. The repository also includes a README.md with usage instructions and example output. The structure is straightforward: one exploit script and one documentation file.
Products and vendors Mallory has correlated with this vulnerability. Open in Mallory to drill down to specific CPE configurations and version ranges.
Vendor-confirmed product mapping. Mallory continuously reconciles this list against your asset inventory.
3 sources tracked across advisories and community write-ups. News coverage will land here when it surfaces.
No news coverage yet. Advisories and community discussion only.
Query your assets running an affected version, and investigate the blast radius.
Every observed campaign linking this CVE to a named adversary.
Malware families riding this exploit, with evidence and IOCs.
YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Cross-references every affected SKU, including bundled OEM variants.
Community discussion across Reddit, Mastodon, and other social sources.