Regulating the Agency of LLM-based Agents

Seán Boddy, Joshua Joseph

As increasingly capable large language model (LLM)-based agents are developed, the potential harms caused by misalignment and loss of control grow correspondingly severe. To addres…