NGINX2Root: Critical NGINX vulnerabilities

Two critical vulnerabilities in NGINX could allow unauthenticated attackers to execute code on affected web servers. Researchers have also demonstrated how a NGINX compromise could be combined with a separate Linux kernel vulnerability to gain root access to the host.

Two critical vulnerabilities in NGINX could allow unauthenticated attackers to execute code on affected web servers. Researchers have also demonstrated how a NGINX compromise could be combined with a separate Linux kernel vulnerability to gain root access to the host.

Disclosed on 18th November 2026, the vulnerabilities highlight the importance of securing the infrastructure supporting websites and APIs, not just the applications themselves.

What are the vulnerabilities?

CVE-2026-123007 and CVE-2026-123633 are two critical vulnerabilities affecting how NGINX handles HTTP/2 traffic. Both are rated Critical by F5, with a CVSS v4.0 score of 9.2.

They form part of a wider group of five NGINX vulnerabilities referred to by some security vendors as NGINX2Root.

  • CVE-2026-123007: A flaw in HTTP/2 flow control that can lead to memory corruption and remote code execution.
  • CVE-2026-123633: A separate HTTP/2 use-after-free vulnerability affecting NGINX deployments that proxy requests to backend servers under specific connection conditions.

CVE-2026-123007 is particularly significant because it affects NGINX versions dating back to 1.9.5 and requires HTTP/2 to be enabled. Researchers have demonstrated exploitation against servers delivering static content, without requiring a vulnerable application behind them.

There are currently no confirmed reports of exploitation in the wild. However, proof-of-concept code for CVE-2026-123007 is publicly available, although it is tuned to specific builds and requires a very large volume of traffic.

Who could be affected?

  1. CVE-2026-123007 affects NGINX versions 1.9.5 to 1.31.8 where HTTP/2 is enabled over HTTPS.
  2. CVE-2026-123633 affects versions 1.19.7 to 1.31.8 where NGINX accepts HTTP/2 traffic and passes requests to a backend, a common reverse-proxy configuration.

Although HTTP/2 is not enabled in the default NGINX configuration, it is widely used across HTTPS websites and enabled by default in some products built on NGINX. Organisations should also consider NGINX included within appliances, container images and other third-party products.

What should organisations do?

F5 released security updates on 18 November 2026. The fixed versions addressing both vulnerabilities are:

  • NGINX Open-Source stable: 1.30.7
  • NGINX Open Source mainline: 1.31.9
  • NGINX Plus: 37.0.8.1 or 37.3.0.1

Where NGINX is supplied through an operating system or another vendor, check its security advisory. Some suppliers backport fixes without changing the upstream version number. If immediate patching isn’t possible, temporarily disabling HTTP/2 can mitigate these two vulnerabilities, although it does not address the other vulnerabilities in the wider release. NGINX must also be fully restarted following an update to ensure the patched executable is running.

How can AppCheck help?

Both CVEs are included in AppCheck’s vulnerability database and identified through passive detection during web application scans.

This works by matching the NGINX version disclosed in server responses to known vulnerable versions.

It’s important to understand the limitations: this identifies potential exposure, not confirmed exploitability. It does not establish whether HTTP/2 is enabled, and detection may not be possible where version information is hidden. Operating system security updates can also be backported while retaining older version numbers.

To review detected NGINX instances across your scans, search for “nginx” in Vulnerability Management. We’re also working on an active detection as a follow-up. However, if you know you’re running an affected version with HTTP/2 enabled, prioritise patching rather than waiting for another scan.

If you have questions or would like help interpreting your results, please get in touch with our team.

Further reading:

Get started with Appcheck

No software to download or install.

Contact us or call us 0113 887 8380

About AppCheck

AppCheck is a software security vendor based in the UK, offering a leading security scanning platform that automates the discovery of security flaws within organisations websites, applications, network and cloud infrastructure. AppCheck are authorised by the Common Vulnerabilities and Exposures (CVE) Program as a CVE Numbering Authority (CNA)

No software to download or install.
Contact us or call us 0113 887 8380