AI SECURITY • 352 / 397
Understand how untrusted text, tools and autonomy create new attack paths.
Knowledge Base Poisoning
Knowledge-base poisoning corrupts the trusted source material an AI system relies on.
Think of it like
Think of Knowledge Base Poisoning as an input-trust or privilege problem: words can influence software decisions, so boundaries must be explicit.
Real life
Email, documents and web pages can become hostile inputs to AI systems through attacks involving Knowledge Base Poisoning.
SRE lens
Compromise of an internal wiki can indirectly compromise RAG answers.
Remember thisKnowledge-base poisoning corrupts the trusted source material an AI system relies on.