Critical Pre-Auth RCE Found in NGINX HTTP/2

Original: Pre-auth RCE in Nginx HTTP/2 handling

Why This Matters

NGINX fronts a huge share of the internet; pre-auth RCE at this scale is rare and severe.

Wiz Research and four other teams independently found two critical unauthenticated RCE flaws in NGINX (CVE-2026-123633, CVE-2026-123007), both CVSS 9.2, affecting 225M+ websites. F5 patched both on Nov 18, 2026.

Two critical remote code execution vulnerabilities in NGINX — the world's most widely deployed open-source web server, powering over 225 million websites — were independently discovered by researchers at Wiz, Ikotas Labs, Metnir, Rapid7, and Xint. Both CVE-2026-123633 (a use-after-free) and CVE-2026-123007 (an integer underflow) carry a CVSS 4.0 score of 9.2 and require HTTP/2 to be enabled on the target server.

Exploitation requires no authentication. A successful attack can give an attacker full code execution on the NGINX host, enabling data theft, service disruption, or website tampering. Because NGINX typically sits in front of multiple backend services, one compromised instance can expose an entire stack — even if the application code behind it is clean.

CVE-2026-123007 traces back to NGINX 1.9.5, released in 2015 — over a decade of exposure. Wiz tested it achieving code execution on a server serving static files, with no backend app or third-party modules required. Wiz reported CVE-2026-123633 to F5 on October 21 with a confirmed ASLR bypass; F5 noted another team had already filed the same bug. The second flaw was reported November 16 and patched within two days. F5 shipped fixes for both on November 18 in a coordinated release. Organizations should update NGINX to the latest release immediately.

Source

wiz.io — Read original →