LLM04:2025: Data and Model Poisoning

Malicious manipulation of training data, fine-tune data, RAG corpus, or embeddings to influence model output, install backdoors, or skew behavior on attacker-chosen triggers.

Found by reviewing RAG ingestion and fine-tuning pipelines for unvalidated or unverified data sources that could bias or backdoor model behavior.

LLM04:2025 maps to CWE-345.

TurboPentest's agentic pentest is powerful and covers a broad range of issues automatically. This particular class is best confirmed in a manual IntegSec engagement, where human pentesters apply deeper methodology and a larger context window than any automated pass.