Who Authorized That? The Delegation Problem in Multi-Agent AI

This article discusses the delegation problem in multi-agent AI, where agents make decisions and access internal systems on a user's behalf. This raises critical security and authorization questions that security teams struggle to answer.