Pre-auth RCE in Nginx HTTP/2 handling
First reported by Wiz ·
If you use NGINX with HTTP/2 enabled, your servers are vulnerable to unauthenticated remote code execution unless patched.
Two critical unauthenticated remote code execution vulnerabilities, CVE-2026-123633 and CVE-2026-123007, have been discovered in NGINX, the widely used web server powering over 225 million websites. These flaws, affecting how NGINX handles HTTP/2 traffic, can allow attackers to take full control of servers, steal sensitive data, tamper with websites, or disrupt services. The vulnerabilities were independently found by multiple research teams, including Wiz Research, Ikotas Labs, Metnir, Rapid7, and Xint, and reported to F5, the company behind NGINX. CVE-2026-123007 has existed since NGINX version 1.9.5 in 2015, while CVE-2026-123633 dates back to around six years ago. Both are rated Critical with a CVSS 4.0 score of 9.2 and require HTTP/2 to be enabled. F5 has released security updates addressing these issues, and organizations are urged to upgrade their NGINX deployments or temporarily disable HTTP/2 if immediate upgrades are not possible.
The discovery of these critical vulnerabilities, particularly the long-standing nature of CVE-2026-123007 dating back to 2015, highlights the persistent risks within widely adopted open-source software. The fact that multiple independent teams, aided by AI tools, found these flaws simultaneously suggests a broader trend: complexity is no longer a sufficient defense against vulnerability discovery. This implies that even deeply embedded and long-standing software components may harbor previously undetectable critical security issues.
For security teams, this accelerates the threat landscape, shortening the window between vulnerability disclosure and active exploitation. The ability of AI to assist in uncovering complex memory-safety issues means that the rate of discovery for such previously elusive flaws is likely to increase. Organizations must adapt their security postures to account for this accelerated discovery cycle, focusing on rapid patching and robust detection mechanisms to mitigate emergent threats.
AI-written summary. May contain errors.