Post

Model Confusion - Weaponizing ML models for red teams and bounty hunters

This DEFCON31 talk write-up details supply chain attack techniques targeting ML pipelines via Hugging Face, analogous to Alex Birsan’s dependency confusion but applied to ML models. Three primary vectors are covered: Model Confusion (publishing malicious models under names that shadow legitimate ones), Organization Confusion (tricking engineers into joining attacker-controlled orgs to gain read/write on their repositories), and typosquatting/watering holes. The post includes a PoC for injecting C2 implants directly into Keras/TensorFlow model architectures (leveraging pickle deserialization and protobuf formats), and notes that ML environments are high-value targets because they run with privileged access to sensitive training data, lack mature supply chain tooling (no Snyk/Artifactory equivalent), and have low detection probability. Detection evasion is compounded by the large binary format of model files making VirusTotal and static analysis difficult.

Read original article

This post is licensed under CC BY 4.0 by the author.