Fetching the paper…

Hacking Back the AI-Hacker: Prompt Injection as a Defense Against LLM-driven Cyberattacks · Around