LiteLLM, an AI tool with 95 million monthly installs, hit by supply chain poisoning; user data security at risk

LiteLLM, an open-source AI API gateway with 95 million monthly installations, recently fell victim to a supply chain poisoning incident, sparking global developer concern. As a core tool powering thousands of enterprises' AI architectures, this attack poses a potential threat to businesses' data security worldwide.

Hacker group TeamPCP launched a three-stage attack, releasing malicious versions (1.82.7 and 1.82.8) on PyPI. These versions steal user credentials—including SSH keys, cloud platform authentication, and cryptocurrency wallets—with highly covert code that infects environments without user interaction.

Security experts recommend: Check your LiteLLM version immediately, look for malicious files, and rotate all cloud keys; downgrade to the secure 1.82.6 version, and audit CI/CD pipelines from the past 48 hours to eliminate any lingering backdoors.