A critical flaw numbered CVE-2025-43859 has been discovered in h11, which is a Python I/O-agnostic, very light-weight HTTP library. With a CVSS score of 9.1, the bug allows for HTTP request smuggling if h11 is used in conjunction with a poorly configured or malfunctioning reverse proxy. The issue lies with h11's flawed handling of line terminators inside chunked transfer encoding, which it processed any two ending bytes instead of the correct CRLF sequence. Such leniency in parsing, which was fixed in h11 version 0.15.0, can lead to critical security issues if misused with some network configurations. The flaw is insecure when h11 and an HTTP proxy disagree about interpreting incoming HTTP messages, specifically in the context of chunked encoding. If a proxy uses naive "read until end of line" semantics, it can misunderstand message boundaries. This disparity allows an attacker to smuggle in a second request that may have sensitive headers (e.g., session cookies) intended for a different environment. As noted in the advisory, the attack would enable an attacker to steal or leak credentials by making the backend server process multiple requests as a single request, violating request isolation and access controls. Developers using h11 in HTTP proxy environments are highly advised to upgrade to version 0.15.0 as soon as possible. It is also important to examine any reverse proxies in the deployment pipeline to ensure that they correctly parse and validate chunked HTTP messages. Failing to apply this patch may expose applications to session hijacking, credential leakage, or unauthorized access to secure endpoints.
A critical security flaw in Oracle WebLogic Server has rapidly become a prime target for attackers worldwide. Identified as CVE-2026-21962, the issue carries the highest possible s...
A new Windows malware called ResokerRAT has been discovered, which allows attackers to secretly control infected systems. This malware uses Telegram instead of traditional servers ...
A vulnerability has been identified in Google Cloud’s Vertex AI platform that could allow unauthorized users to access sensitive data associated with machine learning workloads. ...