DefCamp Day 2 Track #3 - Balcescu 44m talk 6m read Lets make the LLM a Defector: Deceiving LLM into Attacking its Own Agent through natural language Eric Kim, Chunghyun Kang ai securityoffensive security
AI