
Artificial intelligence-assisted security research has uncovered a new set of critical vulnerabilities in NGINX, one of the most widely deployed technologies powering websites and internet applications. The flaws could allow unauthenticated attackers to execute malicious code remotely on vulnerable servers, potentially exposing sensitive information and putting important web infrastructure at risk.
F5, the company responsible for NGINX, has released an out-of-band security update addressing 10 vulnerabilities, including two critical memory-safety flaws rated 9.2 out of 10. Both high-severity vulnerabilities affect the software’s handling of HTTP/2 traffic and could potentially be exploited without requiring an attacker to authenticate first.
The affected technology is used across web servers, cloud environments and application delivery infrastructure. With millions of websites relying on NGINX, the vulnerabilities have attracted attention from security researchers and infrastructure administrators.
F5 recommends that affected users upgrade to NGINX 1.31.9. Although the company has reported no evidence of active exploitation in the wild at the time of its disclosure, the availability of proof-of-concept exploit code makes prompt patching particularly important.
What Is NGINX and Why Does Its Security Matter?
NGINX is an open-source web server and application delivery platform owned by F5. It helps websites and applications receive, process and deliver requests from visitors. Beyond serving web pages, it can also operate as a reverse proxy, load balancer and HTTP cache.
These additional functions make NGINX an important component of modern internet infrastructure. Instead of allowing every incoming request to reach an application’s backend directly, organisations can place NGINX between users and their services to manage traffic, distribute workloads and improve performance.
For example, a website experiencing thousands of simultaneous visitors can use NGINX to distribute incoming requests across multiple application servers. Its caching capabilities can also reduce repeated requests to backend systems, helping websites deliver content more efficiently.
NGINX is used by organisations of different sizes, from smaller websites to large-scale cloud services. Its widespread deployment means that vulnerabilities in its underlying code can have consequences beyond individual websites.
Estimates cited in the original reporting illustrate its scale. WebAtla’s November 2026 figures put NGINX at approximately 73 million websites, representing around 22% of all websites. Netcraft’s figures differ, estimating approximately 225 million websites and 15.3% of the sites it surveyed.
Although these estimates use different measurement approaches, both indicate that NGINX occupies a substantial position in the web infrastructure ecosystem.
Two Critical NGINX Vulnerabilities Could Enable Remote Code Execution
F5’s security notification addresses 10 vulnerabilities, but two stand out because of their potential impact. The flaws are identified as CVE-2026-123633 and CVE-2026-123007, and both carry a reported severity score of 9.2 out of 10.
Remote code execution is among the most serious categories of software vulnerability. If successfully exploited, it can allow an attacker to make a vulnerable system execute commands or code without having legitimate access to the server.
In a web server environment, this could potentially provide an attacker with a route to interfere with applications, access sensitive information or establish further control over the affected infrastructure. The actual consequences would depend on the server’s configuration, privileges and surrounding security controls.
Both vulnerabilities involve memory-safety problems in the way NGINX processes HTTP/2 traffic. They also share an important condition: HTTP/2 must be enabled on the affected server for the described attack scenarios to apply.
CVE-2026-123633: A Use-After-Free Memory Vulnerability
The first critical vulnerability, CVE-2026-123633, involves a memory-management problem known as a use-after-free condition.
This type of vulnerability occurs when software continues attempting to access a section of memory after that memory has already been released. If another operation changes or reuses the same memory region, the application may behave unpredictably and potentially expose an attacker to opportunities for exploitation.
According to the explanation provided by security researcher Tomer Schwartz, the NGINX issue occurs when a connection’s memory is freed while a request still requires access to it.
Memory-safety vulnerabilities can be particularly dangerous in software that processes data from remote users. Web servers regularly handle requests from browsers, applications and automated clients, making correct memory management essential for maintaining security and stability.
CVE-2026-123007: An HTTP/2 Flow-Control Accounting Error
The second critical vulnerability, CVE-2026-123007, involves incorrect flow-control accounting that can result in an invalid memory copy.
HTTP/2 uses flow control to manage how much data is transmitted between communicating endpoints. This helps prevent one side of a connection from overwhelming the other with excessive data.
If software incorrectly tracks these limits or handles the associated memory operations, it can create conditions in which data is copied into an invalid memory location or processed incorrectly.
In this case, researchers identified a flaw in NGINX’s handling of HTTP/2 traffic. The vulnerability is particularly concerning because the affected software versions reportedly extend as far back as NGINX 1.9.5, demonstrating that the underlying issue can affect older deployments as well.
Both critical vulnerabilities were assigned a 9.2 severity rating in the reported disclosure. Administrators should therefore check their installed versions and configuration rather than assuming that an older installation is unaffected.
Researchers Used AI to Discover Longstanding Security Problems
The NGINX vulnerabilities are part of a broader series of discoveries in widely used software that have drawn attention to artificial intelligence-assisted security research.
According to reporting by The Register, Tomer Schwartz, CTO of exposure management and application security at Wiz, worked alongside independent researchers from Ikotas Labs, Metnir, Rapid7 and Xint in identifying the NGINX issues. AI tools were used as part of the research process.
Schwartz described the growing number of AI-assisted vulnerability discoveries as evidence of a changing security landscape. His central warning was that internet-connected systems should be treated as potentially vulnerable because AI is changing both the range of weaknesses researchers can identify and the speed at which longstanding problems can be uncovered.
This development matters because many foundational software projects have existed for years or even decades. Their age and widespread use do not necessarily mean that every security weakness has already been discovered.
AI-assisted analysis can help researchers examine large codebases, identify suspicious memory operations and investigate complex execution paths. However, discovering a potential vulnerability is only one part of the process. Researchers must still validate the issue, understand its impact and develop reliable methods to reproduce it.
The NGINX disclosure demonstrates how this research process can bring attention to problems in software that sits at the centre of everyday internet activity.
NGINX Is Not the First Major Technology Hit by AI-Assisted Vulnerability Research
The latest discovery follows other reports involving critical vulnerabilities in widely used software components. These incidents have raised questions about how organisations should prepare for a period in which automated tools can help uncover weaknesses that may have remained unnoticed for years.
PHP Vulnerabilities and the PHP2Shell Name
In late October 2026, researchers identified two critical vulnerabilities in the PHP runtime. The reported issues could allow unauthenticated attackers to execute code remotely on vulnerable web servers or access sensitive information stored in memory.
The flaws were subsequently associated with the name PHP2Shell. The discovery highlighted the potential consequences of vulnerabilities in software that powers a large number of websites and web applications.
Apache HTTP Server Vulnerability
Apache HTTP Server was also included in the wider series of reported security discoveries, with attention drawn to CVE-2026-119618.
Apache is another foundational web server technology, and vulnerabilities affecting it can have implications for organisations that depend on it to deliver websites and applications.
Java Memory-Corruption Vulnerability
Another reported discovery involved a critical memory-corruption vulnerability in Java’s Base64 decoder, referred to as Java4Shell.
The issue was significant enough to attract attention from the US Cybersecurity and Infrastructure Security Agency (CISA), which warned organisations about the need to address the vulnerability.
Together, these incidents illustrate why security teams cannot rely exclusively on a software project’s reputation or age when evaluating risk. Even established technologies require continuous security testing, maintenance and timely patching.
Rapid7 Publishes Proof-of-Concept Exploit for an NGINX Vulnerability
One development that makes the NGINX disclosure particularly important is the availability of proof-of-concept exploit code.
Rapid7 reportedly released a proof-of-concept demonstrating how CVE-2026-123007 could be used to achieve unauthenticated remote code execution against a vulnerable Linux/amd64 Docker image running NGINX version 1.31.8.
A proof of concept, commonly shortened to PoC, is a demonstration that a vulnerability can be triggered under specific conditions. It does not automatically mean that attackers are exploiting the weakness against real-world systems, but it can make the technical risk easier to understand and reproduce.
For defenders, publicly available exploit demonstrations can be useful when testing whether systems are vulnerable and validating security controls. However, they also reduce the amount of work required for other researchers or malicious actors to investigate an exposed weakness.
Organisations should therefore avoid delaying security updates simply because there is no confirmed evidence of widespread exploitation.
Has Anyone Exploited the NGINX Vulnerabilities?
Despite the seriousness of the flaws, F5 said it had not found evidence of active exploitation in the wild at the time of its disclosure.
An F5 spokesperson told The Register that the company was continuing to monitor the situation and would share additional information through its myF5 security communication channels as more details became available.
This distinction is important. A critical vulnerability and a working proof-of-concept exploit indicate that a system may be at serious risk, but they do not establish that attackers have already compromised production servers.
Organisations should also recognise that the absence of publicly confirmed exploitation is not proof that every vulnerable system is safe. Internet-facing infrastructure can be exposed to scanning and attack attempts, and security teams may not immediately identify suspicious activity.
Administrators should use the disclosure as an opportunity to review their NGINX deployments, verify whether HTTP/2 is enabled and apply the recommended security update.
How to Protect NGINX Servers Against the Critical Vulnerabilities
F5’s recommended response is to update affected NGINX installations to version 1.31.9. Organisations should prioritise systems exposed to the internet, particularly those with HTTP/2 enabled.
Security teams should follow a structured approach when responding to the disclosure.
1. Identify the Installed NGINX Version
Start by checking the NGINX version running on every relevant server, container and cloud environment. Do not limit the assessment to production web servers. Development systems, staging environments and internal services may also contain vulnerable installations.
Organisations with large infrastructure deployments should use configuration management and vulnerability scanning tools to identify affected instances consistently.
2. Upgrade to NGINX 1.31.9
Update affected installations to NGINX 1.31.9 in accordance with F5’s security guidance. Administrators should obtain packages and updates from trusted distribution channels and follow the appropriate deployment instructions for their environment.
For organisations using vendor-maintained packages or container images, check the relevant provider’s security advisories to confirm which patched release applies to their specific distribution.
3. Check HTTP/2 Configuration
Both highlighted critical vulnerabilities require HTTP/2 to be enabled under the reported attack conditions. Administrators should review server configuration files and determine where the protocol is active.
Disabling HTTP/2 may reduce exposure to these specific attack paths in some environments, but it should not be treated as a substitute for installing the security update. The broader advisory covers 10 vulnerabilities, and administrators should assess all applicable issues.
4. Review Internet-Facing Infrastructure
Prioritise systems that accept requests from the public internet. These servers are more directly exposed to remote attackers and should receive prompt attention.
Review reverse proxies, load balancers, web application gateways and containerised NGINX deployments. A vulnerability in an infrastructure component can affect multiple applications if they depend on the same service.
5. Monitor for Suspicious Activity
After patching, examine server logs and monitoring data for unusual requests, unexpected process behaviour or unexplained changes. Organisations should investigate suspicious activity rather than assuming that installing the update automatically resolves any earlier compromise.
If there are indications of exploitation, follow the organisation’s incident-response procedures and assess whether sensitive data or connected systems may have been affected.
Why AI-Discovered Vulnerabilities Could Change Cybersecurity
The NGINX incident highlights a growing challenge for software developers, security researchers and infrastructure operators. AI-assisted security tools are making it easier to investigate large codebases and identify weaknesses in components that are deeply embedded in modern technology.
This can produce significant benefits. Discovering vulnerabilities before criminals exploit them gives developers an opportunity to fix weaknesses and improve the security of widely used software.
However, faster discovery also places pressure on organisations responsible for maintaining those systems. Security teams may have to investigate and patch multiple critical vulnerabilities across different technologies in a shorter period than traditional processes were designed to handle.
For companies operating large web infrastructures, this means vulnerability management must become a continuous process. Knowing which software is installed, where it is exposed and how quickly updates can be deployed is increasingly important.
It also reinforces the value of defence-in-depth. Timely patching should be supported by restricted service permissions, network monitoring, secure configuration and reliable incident-response procedures.
Final Verdict: NGINX Administrators Should Not Delay This Security Update
The discovery of 10 NGINX vulnerabilities, including two critical flaws that could enable unauthenticated remote code execution, is an important security warning for organisations relying on the software. The fact that the issues affect HTTP/2 processing and include a vulnerability reaching back to older NGINX releases makes checking existing deployments especially important.
AI-assisted research is helping uncover weaknesses in foundational internet technologies, but the immediate priority for administrators is practical: identify affected systems, review HTTP/2 configurations and install the recommended security update.
F5 has reported no evidence of active exploitation at the time of disclosure, but Rapid7’s proof-of-concept demonstration shows why organisations should not wait for confirmed attacks before responding.
Administrators should upgrade affected NGINX installations to version 1.31.9 as soon as possible and verify that their web servers, containers and cloud deployments are protected.
For breaking news and live news updates, like us on Facebook or follow us on Twitter and Instagram. Read more on Latest Technology on thefoxdaily.com.

COMMENTS 0