Signal

Cybersecurity company Xint says it found RCE vulnerabilities in Nginx using GPT 6.1 Sol and GPT-6 Luna; Wiz says it had separately found the same flaws

First reported by Forbes ·

The signal ●●●● Compiled by AI from Forbes, Techmeme, Hacker News and Wiz Blog
Why you might care

If you manage web servers, you must update NGINX to patch critical vulnerabilities that allow unauthenticated remote code execution.

What happened

Cybersecurity firms Xint and Wiz independently discovered two critical remote code execution (RCE) vulnerabilities in NGINX, a widely used web server software. The flaws, tracked as CVE-2026-123633 and CVE-2026-123007, have CVSS 4.0 scores of 9.2 and require HTTP/2 to be enabled. Exploitation could allow unauthenticated attackers to take control of affected NGINX servers, steal sensitive data, or disrupt services. Wiz Research utilized its AI scanning solution, Wiz Atlas, and AI models like Gemini 4 Carbon and ChatGPT 5.6 Sol to identify and validate these vulnerabilities. The disclosure timeline shows independent reporting by multiple teams, including Ikotas Labs, Metnir, and Rapid7, to F5, the company behind NGINX. Wiz first reported CVE-2026-123633 on October 21, 2026, and CVE-2026-123007 on November 16, 2026. F5 has released security updates for both NGINX OSS and NGINX Plus to address these issues.

What it means

The independent discovery of these critical NGINX vulnerabilities by multiple security teams, leveraging AI tools, signals a significant shift in the cybersecurity threat landscape. Historically complex memory-safety flaws, like use-after-free and integer underflows, have been difficult to find and exploit, acting as a de facto barrier. However, advanced AI models are now capable of tracing intricate code interactions, democratizing the discovery of such deep-seated issues for both researchers and potential attackers. This implies that long-standing vulnerabilities in foundational internet infrastructure may be more prevalent than previously understood and could be uncovered with increasing speed and sophistication.

This development directly impacts organizations relying on NGINX for their web services, as their servers are now demonstrably vulnerable to sophisticated, AI-assisted attacks. The fact that these flaws were present for years underscores the potential for numerous undiscovered vulnerabilities in widely used software. Security teams face an accelerated challenge: rapidly identifying and patching these newly exposed risks while simultaneously preparing for a future where AI-driven vulnerability discovery becomes the norm, potentially shortening the window between disclosure and widespread exploitation.

AI-written summary. May contain errors.

AI Amazon