Researchers propose Consilience, a framework for coordinating communication among multi-agent systems with limited information. It ensures that each conversational action is appropriate and provides a guarantee of decision accuracy. In experiments, Consilience improved performance over existing protocols, sometimes even surpassing the baseline where all agents have complete information. --- Why it matters: This matters to AI researchers because it addresses a fundamental challenge in multi-agent systems: how to effectively coordinate communication when each agent has limited information. Consilience's framework and results demonstrate that adaptive communication control can be more valuable than simply increasing available information. Source: https://arxiv.org/abs/2608.20564