AI RESEARCH

On the Reliability Limits of LLM-Based Multi-Agent Planning

arXiv CS.LG

ArXi:2603.26993v1 Announce Type: cross This technical note studies the reliability limits of LLM-based multi-agent planning as a delegated decision problem. We model the LLM-based multi-agent architecture as a finite acyclic decision network in which multiple stages process shared model-context information, communicate through language interfaces with limited capacity, and may invoke human review. We show that, without new exogenous signals, any delegated network is decision-theoretically dominated by a centralized Bayes decision maker with access to the same information.