cyberpress.org·20 november 2026
Attackers Target LiteLLM AI Gateways Using BadHost Authentication Bypass and RCE Vulnerabilities

AI gateways
aiohttp
ANY.RUN
attack surface reduction
authentication bypass
CVE-2026-42271
CVE-2026-48710
Horizon3.ai
LiteLLM
malformed Host headers
patch management
Python
Remote Code Execution
Starlette
WAF rules
cyberpress.org
Attackers are probing LiteLLM AI gateways with malformed Host headers to bypass authentication and run commands on exposed servers. The attacks target CVE-2026-42271 and CVE-2026-48710, which can be chained into unauthenticated remote code execution. Horizon3.ai published a write-up showing how the flaws can be exploited. Defenders should patch Starlette to version 1.0.1 or later, update LiteLLM, and implement WAF rules to mitigate the vulnerability.


