10-Year-Old NGINX HTTP/2 Flaw Lets Unauthenticated Attackers Execute Code on Servers
Our summary
A decade-old vulnerability in NGINX's HTTP/2 implementation, tracked as CVE-2026-123007 and nicknamed "nginx2root," lets an unauthenticated attacker execute code on any server with HTTP/2 enabled. The bug has existed since version 1.9.5 (September 2015) and affects every open-source release up to 1.31.8, as well as NGINX Plus builds from 37.x before the patched versions and R34 through R36. F5 has now shipped fixes in 1.31.9 and 1.30.7, which also resolve four related high-severity issues (CVE-2026-123633, CVE-2026-123634, CVE-2026-121810, CVE-2026-113425) that NGINX scores 9.2 on CVSS 4.0, though researchers from Wiz, Rapid7 and Xint note successful exploitation depends on specific conditions like disabled/bypassable ASLR, and code execution initially only grants NGINX worker-level privileges, with Xint demonstrating a path to full root via a separate Linux kernel SCTP flaw. No in-the-wild exploitation has been confirmed and the flaw isn't yet in CISA's KEV catalog, but public proof-of-concept code already exists, making a prompt upgrade and full restart (not just a reload) important, especially since Ubuntu and Debian haven't yet shipped patched packages.
The Hacker News publishes this story on its own site; we link to it rather than reprint it.
Worried this affects your company?
Discuss a security assessment of your internet-facing systems. Scope agreed before testing.