LiteLLM supply chain attack exposes over 2,500 organizations
Cybercriminals compromised LiteLLM, an open-source tool for connecting applications with AI models, in March 2026 by uploading malicious packages to PyPI for about 40 minutes. CloudSEK identified over 2,500 organizations potentially affected, with approximately 434,000 CI/CD pipelines exposed to sensitive credentials and secrets.
What happened
Cybercriminal group Team PCP compromised LiteLLM in March 2026, injecting malicious code into the open-source tool used to connect applications with AI models.
Malicious versions of LiteLLM were available on PyPI, the Python package repository, for approximately 40 minutes before being removed.
Reported impact
CloudSEK identified over 2,500 organizations potentially exposed, including NVIDIA, Samsung Electronics, Cisco Systems, Siemens, S&P Global, ServiceNow, Deloitte, Vodafone, X Corp, Zscaler, FedEx, Volkswagen, Thales and London Stock Exchange Group.
Approximately 434,000 CI/CD software pipelines were potentially linked to the exposure. Potentially accessible information included cloud credentials, source-code access, server keys, development secrets and AI API keys.
Index analysis
CI/CD systems automatically build, test and deploy software and can download packages without developers manually checking every component, allowing malicious packages to quickly reach large numbers of corporate systems despite brief availability.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.