Skip to content
NewA calmer, faster ZeroHour
Exploit / PoChighExploited in the wild

Attackers scan LiteLLM servers to chain two flaws into RCE

Original title: Hackers Actively Scan LiteLLM AI Servers for Critical Flaws Enabling RCE Attacks

Cyber Security Newsby Tushar Subhra Dutta ()4 min read
Part of a story covered by 2 sourcesScans chain LiteLLM and Starlette flaws for unauthenticated RCEMerged summary & timeline
AI summary · x-ai/grok-4.7Summaries by Vani AI

ISC honeypots saw probes to a LiteLLM test path that tried to run id using a fake token and odd Host headers.

Attackers are scanning for exposed LiteLLM AI gateways and chaining CVE-2026-42271 with CVE-2026-48710 to gain remote code execution without a valid login. Internet Storm Center honeypots received requests to /mcp-rest/test/tools/list that supplied the id command and malformed Host values. CISA added CVE-2026-42271 to the KEV catalog on June 8, 2026, and exploitation was reported in July. LiteLLM 1.83.7 or later and Starlette 1.0.1 or later contain the fixes.

Unconfirmed: Honeypot scans show exploit attempts; the article does not quantify successful compromises.

Key facts
  • CVE-2026-42271 affects LiteLLM from 1.74.2 before 1.83.7 and runs a supplied command.
  • CVE-2026-48710 affects Starlette before 1.0.1 through unchecked Host headers.
  • Wiz found 294 of 3,074 exposed instances accepted a default master key or had no authentication.
  • Scans used Host values such as localhost] and a bearer token of nothing.
Security assessment · AI, from the articleExploited in the wild

Impact: Unauthenticated remote code execution on the LiteLLM host

ProductAffectedFixed in
LiteLLM1.74.2 up to, but not including, 1.83.71.83.7 or later
Starlettebefore 1.0.11.0.1 or later
What to do
  1. 1.Upgrade LiteLLM to 1.83.7 or later and Starlette to 1.0.1 or later.
  2. 2.Remove LiteLLM admin and test endpoints from the public internet.
  3. 3.If a vulnerable server was exposed, rotate the master key and every provider key.
Analyze CVE-2026-42271 with Vani AIImpact, affected versions, exploitation, mitigations and detection

Vulnerabilities mentioned

All
CVEVulnerabilityCVSSEPSSSignalsPublished
CVE-2026-42271
Command Injection in BerriAI LiteLLM AI Gateway Exploited in the Wild
8.791%KEV
KEV
CVE-2026-48710
HTTP Request Smuggling in Kludex Starlette (CVE-2026-48710)
6.570%KEVPoC ×3
KEV

From the original article

4 min read at cybersecuritynews.com

Attackers are now scanning the internet for exposed LiteLLM servers. They want to chain two flaws that give them remote code execution (RCE) without a login. The scans target a test endpoint in the LiteLLM AI gateway and abuse a weak spot in the Starlette web framework that it relies on.

Cyber Security News

By Tushar Subhra Dutta · cybersecuritynews.com

Read the full article at Cyber Security News

A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.