mlsanityData to 14 Oct 2026

Agentic Societies Need a Social Harness

arXiv 2609.17527 ↗First posted 15 Sep 2026cs.MA cs.AI cs.NIcited by 0

On the map: LLM agents and tool use → Social simulation and multi-agent debate

Abstract (arXiv)

An agentic society is a collection of AI agents that coordinate autonomously across trust boundaries, on behalf of different principals whose objectives may only partially align. We show experimentally that in agentic societies even honest, competent agents often fail to reach satisfactory outcomes with existing harnesses and messaging primitives, and that faulty or malicious agents can stall collaboration, influence outcomes, and pursue other harmful goals by exploiting vulnerabilities in communication (``speech''). We argue that agentic societies need a social harness for inter-agent interactions, in addition to each agent's personal harness, which manages its private context and communication with its principal. We propose a layered architecture for social harnesses which (i) prevents classes of failures outright, (ii) enables agents to detect invalid messages at runtime, and (iii) supports post-facto investigation and consequences, and highlight directions for future research to realize these capabilities.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Social simulation and multi-agent debate, one of 16 directions in LLM agents and tool use. Builds on 16 of its 47 references that are in the map. First posted on arXiv 15 Sep 2026.

Builds on

16 of 47 references.

Closest ancestors · 1

Same direction, newest first

  1. 2026Can AI Agents Agree?1 Mar

Other references · 15

  1. 2023AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversationcited by 1,988
  2. 2025Why Do Multi-Agent LLM Systems Fail?cited by 604
  3. 2024Agent Security Bench (ASB): Formalizing and Benchmarking Attacks and Defenses in LLM-based Agentscited by 369
  4. 2025Defeating Prompt Injections by Designcited by 297
  5. 2025MultiAgentBench: Evaluating the Collaboration and Competition of LLM agentscited by 178
  6. 2024RE-Bench: Evaluating frontier AI R&D capabilities of language model agents against human expertscited by 175
  7. 2025GDPval: Evaluating AI Model Performance on Real-World Economically Valuable Taskscited by 158
  8. 2024Prompt Infection: LLM-to-LLM Prompt Injection within Multi-Agent Systemscited by 133
  9. 2025Red-Teaming LLM Multi-Agent Systems via Communication Attackscited by 122
  10. 2024Internet of Agents: Weaving a Web of Heterogeneous Agents for Collaborative Intelligencecited by 86
  11. 2024PsySafe: A Comprehensive Framework for Psychological-based Attack, Defense, and Evaluation of Multi-agent System Safetycited by 80
  12. 2025G-Safeguard: A Topology-Guided Security Lens and Treatment on LLM-based Multi-agent Systemscited by 59
3 more
  1. 2026Agents of Chaoscited by 32
  2. 2026AgentWorm: Self-Propagating Attacks Across LLM Agent Ecosystemscited by 17
  3. 2025Byzantine-Robust Decentralized Coordination of LLM Agentscited by 14

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Mechanism Design Is Not Enough: Prosocial Agents for Cooperative AI
  2. 2026Conformity and Social Impact on AI Agents
  3. 2026Talk, Judge, Cooperate: Gossip-Driven Indirect Reciprocity in Self-Interested LLM Agents
  4. 2025The Traitors: Deception and Trust in Multi-Agent Language Model Simulations
  5. 2025Revisiting Gossip Protocols: A Vision for Emergent Coordination in Agentic Multi-Agent Systems
  6. 2026Deception and Communication in Autonomous Multi-Agent Systems: An Experimental Study with Among Us
  7. 2026Less Interaction But More Explanation: A Communication Perspective on Agentic AI Interfaces
  8. 2026Governance by Design: A Parsonian Institutional Architecture for Internet-Wide Agent Societies