LiteLLM compromised in supply chain attack stealing credentials
The popular LiteLLM tool, an AI gateway that connects applications to multiple AI model providers, was distributed in a compromised version 1.107.0 through PyPI and Docker images after attackers gained access to three team members' GitHub accounts using stolen OAuth tokens, credentials that let an application act on a user's behalf without knowing their password. The malicious release automatically installed two programs: one that harvested cloud keys, private keys and browser data, and a second that checked a GitHub account every hour for new commands from the attacker. LiteLLM pulled the compromised version and revoked all affected tokens, but has not yet determined how the tokens were obtained. For any business using such gateways to connect its own tools to Claude, ChatGPT or other models, the incident is a reminder that AI infrastructure inherits the same supply chain risks as any other open-source software.
Sources
This summary was written automatically with AI from the public sources listed above and checked by an independent AI reviewer before publication. For full details, read the original source.