## LiteLLM v1.83.0 Patches Critical Auth Bypass Flaw (CVE-2026-35029), Prevents RCE
A critical security vulnerability in the LiteLLM proxy server has been patched, exposing authenticated users to a path for remote code execution. The flaw, tracked as CVE-2026-35029, resided in the `/config/update` endpoint, which failed to enforce admin role authorization. This oversight created a direct channel for an attacker to seize control of the system.

Any user already authenticated to the platform could exploit this endpoint to perform unauthorized administrative actions. The primary risks included the ability to modify critical proxy configurations and environment variables. More alarmingly, an attacker could register custom pass-through endpoint handlers that point to malicious, attacker-controlled Python code, effectively achieving remote code execution on the host server. The vulnerability also allowed for the arbitrary reading of sensitive data.

The patch is delivered in LiteLLM version 1.83.0, which upgrades from the vulnerable 1.82.4. This update is marked as a security fix, and the associated GitHub Security Advisory (GHSA-53mr-6c8q-9789) provides full details. Organizations and developers using LiteLLM must immediately apply this update to close the authorization gap and prevent potential system compromise and data exfiltration.
---
- **Source**: GitHub Issues
- **Sector**: The Lab
- **Tags**: vulnerability, CVE-2026-35029, remote code execution, auth bypass, AI proxy
- **Credibility**: unverified
- **Published**: 2026-04-08 19:27:30
- **ID**: 55574
- **URL**: https://whisperx.ai/en/intel/55574