Agentic Societies Need a Social Harness — New arXiv Paper on Inter-Agent AI Coordination

Summary

A new arXiv paper (2609.17527) by Tapan Chugh, Vidushi Singh, Krish Jain, Arvind Krishnamurthy, and Ratul Mahajan argues that agentic societies — collections of AI agents coordinating autonomously across trust boundaries on behalf of principals with partially aligned objectives — require a dedicated social harness for inter-agent interactions. The authors show experimentally that even honest, competent agents often fail to reach satisfactory outcomes with existing harnesses and messaging primitives, while faulty or malicious agents can stall collaboration, manipulate outcomes, or pursue harmful goals by exploiting vulnerabilities in agent-to-agent communication ('speech'). The paper distinguishes each agent's personal harness, which manages private context and communication with its principal, from the proposed social harness. They present a layered architecture that (i) fundamentally prevents multiple classes of failures, (ii) enables agents to detect ineffective messages at runtime, and (iii) supports post-hoc investigation and accountability. The paper also outlines future research directions toward realizing these capabilities.