20 Prompt Injection Techniques Every Red Teamer Should Test | by Facundo Fernandez | Medium

Prompt injection is the SQLi of the LLM era. Every organization rushing to adopt agentic apps, copilots, and AI integrations is exposed to this attack surface. Red teamers need a systematic playbook to evaluate vulnerabilities, not just a few copy-pasted jailbreaks. Here’s a breakdown of 20 techniques across 7 categories, with explanations and very-high-level example payloads, that every red team operator should test on engagements. 2. Role Reversal Attack: Convince the model it is the “system” or “developer” and therefore above restrictions. Example: 3. Chain-of-Thought Extraction Attack: Force the model to reveal its hidden reasoning or instructions. Example: 4. Base64 Encoding Attack: Encode payloads to bypass filters. Example: 5. Unicode/Whitespace Abuse Attack: Use homoglyphs and zero-width spaces to sneak in restricted text. Example: 6. Nested Escapes Attack: Hide payloads inside formats the model must parse. Example: 7. Do-Anything-Now (DAN) Style Attack: Create a fictional unre

saved by

• CLL