Hacking Back the AI-Hacker: Prompt Injection as a Defense for LLM-Attackersarxiv.org2 points·emk_709··0 commentsOpen articleSaveView on HN