Skip to content
Your guide to common prompt defenses
bugcrowd.com | blog | #ai-security | #prompt-injection | #llm | #prompt-defense | #input-sanitization | #adversarial-inputs
Summary
This guide reviews common LLM prompt defenses—input preprocessing, sanitization, blocklists, and allowlists—and explains why non-deterministic model behavior lets attackers bypass them.
- Published
- Collected