Breaking RAG Systems - Exploiting Vulnerabilities and Hardening Your GenAI Applications

Explore critical security vulnerabilities in Retrieval Augmented Generation (RAG) systems through this hands-on conference talk that demonstrates real-world attacks and defensive strategies. Learn how RAG systems, while becoming essential for enterprise GenAI applications, introduce unique security risks that most development teams overlook. Witness live demonstrations of three major attack vectors: hallucination injection techniques that force models to confidently return false information, prompt manipulation methods that bypass business logic restrictions, and vector database poisoning attacks that compromise RAG retrieval results. Discover step-by-step remediation strategies including securing vector databases against poisoning attacks, building multi-stage guardrails to catch manipulated inputs, and implementing robust retrieval validation techniques. Gain practical code patterns and configurations that can be immediately applied to strengthen your own RAG applications against these emerging security threats.