AI agents found a zero-day on their own
Attackers used the open-source tool Cairn
Security
F5 patches HTTP/2 code-execution bugs · found by five teams
Written with AI assistance, reviewed by the ACN desk.
“The flaws were already there; AI helped us find them.”
F5 has released security updates for critical flaws in NGINX, the open-source web server. The flaws let an attacker run code on a server without logging in. Wiz Research’s write-up says it found two of them with Wiz Atlas, its AI code scanning harness.
Wiz rates both flaws critical, with a CVSS 4.0 score of 9.2. Both require HTTP/2, a newer web protocol, to be enabled. One is a use-after-free bug, a type of memory error. The other is an integer underflow in HTTP/2 flow control. Fixed releases include NGINX 1.30.7 and 1.31.9. F5 released the updates on November 18.
Wiz says Ikotas Labs, Metnir, Rapid7 and Xint also found and reported the flaws independently. Xint’s report lists five NGINX flaws in total. It says a researcher found the first while preparing for ZeroDay Cloud 2026. Xint says a fully automated scan then found the other four without human guidance. Xint writes that “AI-assisted research is finding bugs faster than ever, even in code as heavily studied as nginx.”
Wiz says NGINX powers more than 225 million websites. It advises upgrading, or turning HTTP/2 off until a fix is installed. Xint warns that distribution packages may carry fixes without changing the reported version number.
“The flaws were already there; AI helped us find them.”
Sponsor
Spot an error? Email corrections@aicodingnews.si. Fixes are logged on the corrections page.
Sponsor
Sponsor
Attackers used the open-source tool Cairn
CVE-2026-114938 · Oracle has released a patch
High- and critical-severity bugs · from January 1, 2027
Google Threat Intelligence Group · first with no human in the loop
Sponsor
Sponsor