Critical RCE in Wp2shell Could Fetch $500,000 on Exploit Markets

Key Takeaways

• An advanced AI model, GPT-5.6 Sol Ultra, identified a critical pre-authentication RCE vulnerability in WordPress.
• The exploit chain allows unauthenticated attackers to gain administrator privileges and execute arbitrary code.
• The vulnerability leverages flaws in WordPress’s Batch API, REST API parameter sanitization, and changeset handling.
• The potential value of such an exploit on the black market is estimated at up to $500,000.
• WordPress administrators should apply all available security updates immediately and monitor for suspicious activity.

A cutting-edge artificial intelligence, GPT-5.6 Sol Ultra, has reportedly uncovered a severe pre-authentication remote code execution (RCE) vulnerability within the WordPress content management system. This discovery, made after an estimated $25 worth of AI processing power, underscores the transformative potential of advanced AI models in the field of cybersecurity vulnerability research.

Researchers at Searchlight Cyber engaged GPT-5.6 Sol Ultra to conduct an audit of a local WordPress source code instance. Utilizing four distinct AI agents over a period exceeding six hours, the model successfully pinpointed a comprehensive exploit chain. This chain, if leveraged, could enable an unauthenticated malicious actor to escalate privileges to administrative levels and ultimately achieve arbitrary code execution on affected WordPress installations.

The financial implications of such a vulnerability are substantial. Security experts estimate the market value of a reliable zero-click or pre-authentication RCE affecting widely deployed software like WordPress to be as high as $500,000, reflecting the severe potential damage and broad impact of such an exploit.

Understanding the Exploit Chain

The Initial Vector: WordPress Batch API

Given that WordPress powers over half a billion websites globally, any RCE affecting its default configuration carries immense significance. The reported attack sequence initiates within WordPress’s Batch API, which is accessible via the /wp-json/batch/v1 endpoint. This API is designed to consolidate multiple REST requests into a single HTTP request for efficiency.

The core of the vulnerability lies in a reported desynchronization between validation results and the corresponding REST handlers, exacerbated by an error-handling condition. This flaw could permit an attacker to successfully validate one request while its execution is redirected against a different endpoint handler, effectively bypassing the parameter sanitization mechanisms intended for REST API routes.

Researchers specifically identified a combined issue within the posts endpoint. Here, an author_exclude value, when supplied as a scalar instead of the expected integer array, could be passed to a database query without proper escaping. To circumvent the Batch API’s usual restriction on GET requests, the AI model devised a nested batch request. This technique bypassed method validation, subsequently allowing for the injection of SQL-related parameters without scrutiny.

Escalation to RCE

This initial maneuver resulted in a pre-authentication SQL injection vulnerability, enabling the attacker to read sensitive data from the WordPress database. However, escalating from SQL injection to full remote code execution required a more intricate series of steps.

The Searchlight Cyber team detailed how the AI model exploited WordPress’s in-memory post cache and the oEmbed caching mechanism. This allowed for the creation of controlled, database-backed post records, which were then converted into specialized customize_changeset objects.

WordPress changesets are critical components that facilitate configuration modifications on a site under the authority of a specified user. By forging a changeset linked to administrator user ID 111, the exploit could temporarily trick WordPress into granting administrator privileges during the request processing phase.

The final stage of the attack capitalized on WordPress hooks and a cleverly crafted post hierarchy cycle. A forged post compelled WordPress to invoke the parse_request hook, which replayed the malicious batch request. Crucially, during this second pass, the system still retained the temporarily elevated administrator privileges. This window allowed the attacker to create a new administrator account and subsequently install a malicious plugin, culminating in complete code execution.

Searchlight Cyber delayed public reporting of this critical vulnerability to provide ample time for defenders to implement necessary patches. Independent cybersecurity researchers, including Calif and Hacktron, successfully reproduced the entire exploit chain before the proof-of-concept code was made publicly available on GitHub.

What You Should Do

• Apply Updates Immediately: Ensure your WordPress installation, themes, and plugins are updated to the latest available versions. This is the most crucial step to mitigate known vulnerabilities.
• Monitor Logs: Regularly review web server and WordPress access logs for suspicious requests targeting the /wp-json/batch/v1 endpoint. Look for unusual parameters or request patterns.
• Utilize Security Scanners: Organizations can check for potential exposure to this and similar vulnerabilities using the hosted scanner at wp2shell.
• Implement Web Application Firewalls (WAFs): A properly configured WAF can help detect and block malicious requests targeting WordPress APIs and other attack vectors.
• Regular Backups: Maintain frequent and secure backups of your WordPress database and files to facilitate rapid recovery in case of a compromise.

Disclaimer: HackersRadar reports on cybersecurity threats and incidents for informational and awareness purposes only. We do not engage in hacking activities, data exfiltration, or the hosting or distribution of stolen or leaked information. All content is based on publicly available sources.