JadePuffer agentic ransomware now encrypts training data and model checkpoints

The JadePuffer operation, documented as the first agentic ransomware campaign, has been upgraded with custom malware that goes after AI assets specifically. The new component, called EncForge, encrypts training datasets, vector databases, and model checkpoints rather than ordinary business files. That shifts the extortion target to the artifacts an AI team cannot easily rebuild and that often sit outside normal backup routines. JadePuffer's original intrusions ran through a critical unauthenticated code execution flaw in Langflow, an open source framework for building AI applications, with an autonomous agent then handling reconnaissance, credential theft, and persistence on its own.

Check

Identify where training datasets, vector databases, and model checkpoints live, confirm they are covered by backups that ransomware cannot reach, and check that Langflow and similar AI orchestration tools are patched.

Affected

Teams running AI workloads with exposed orchestration platforms; the EncForge component encrypts training data, vector databases, and model checkpoints, assets that are expensive to rebuild and often left outside standard backup coverage.

Fix

Back up AI assets with offline or immutable copies, keep orchestration platforms patched and off the public internet, restrict the credentials those workloads hold, and include model stores in recovery testing.