Agentic AI Security Red Teaming LLMs [Local Lab]: Agentic Overdelegation Demo to explore how AI agents can be manipulated to misuse tools Mamta UpadhyayJune 18, 2025June 19, 2025