Google Mandiant Infiltrates TeamPCP Supply-Chain Group from March 2026 Onward

Google Threat Intelligence Group researcher Austin Larsen disclosed at LABScon that a Mandiant persona joined TeamPCP at its formation after building trust with an early member. The analyst observed repeated compromises of Trivy, LiteLLM, TanStack, and Mistral AI infrastructure, plus deployment of the Mini Shai-Hulud worm. Google used the access to issue targeted breach notifications to more than 1,000 affected organizations and to document credential theft from GitHub, OpenAI, and the European Commission.

Australian Federal Police charging documents and the earlier Ars Technica reporting confirm the two Australians made repeated operational security errors that allowed Google to pass identifiers to law enforcement. ShinyHunters, initially partnered with TeamPCP, later supplied additional signals intelligence after the relationship collapsed. These data points align with the pattern of supply-chain campaigns tracked in the 2025 Shai-Hulud incidents and the 2024 XZ Utils compromise.

The infiltration enabled direct disruption of victim exploitation attempts rather than post-breach forensics alone. Mandiant's continuous presence from day one contradicts prior public accounts that portrayed TeamPCP as opaque until the arrests. Future operations against similar small, high-velocity groups will likely prioritize early persona placement over signature-based detection.