Security researchers have identified what they are calling the largest AI supply chain breach uncovered so far in 2026, tied to the widely used LiteLLM project. According to threat intelligence firm CloudSEK, the compromise may have touched more than 2,500 companies and roughly 434,000 CI/CD pipelines worldwide. CloudSEK flagged NVIDIA, Amazon Web Services (AWS), Cisco Systems, Salesforce, Siemens AG, X Corp (Twitter), and Orange S.A. as high-confidence matches in its exposure dataset, though it notes that a “high confidence” match reflects the strength of exposure evidence, not confirmed proof of compromise or data theft. The breach began with a trusted tool inside LiteLLM’s build process: attackers took over the Trivy security scanner using a leaked automation token that had been rotated but never fully revoked, leaving an approximate 20-day window to force-push malicious code over Trivy’s published version tags. The poisoned scanner flowed into LiteLLM’s build system, producing two compromised releases published to the Python Package Index: versions 1.82.7 and 1.82.8. Those packages were live on PyPI for roughly 40 minutes, enough to seed a global exposure event since automated pipelines install dependencies at machine speed. The malicious code executed through a `.pth` file that runs automatically when Python starts, requiring no explicit import of LiteLLM to trigger it.