Norm Enforcement for AI Agents: Robustly Shaping Behavior in Multi-Agent Systems

arXiv cs.AI Papers

Summary

This paper studies norm enforcement mechanisms to shape behavior of language model agents in multi-agent systems. The authors propose robust mechanisms that estimate agent reliability over time and apply escalating penalties to resist exploitation.

arXiv:2607.09766v1 Announce Type: new Abstract: AI agents are increasingly deployed in shared environments where they pursue diverse goals and compete for rewards. This multi-agent competition can lead to behaviors that serve individual gains at collective cost -- for instance, marketing agents may post misleading content as a result of competing for engagement on social media. Human societies address such problems through norms that constrain acceptable behavior, supported by enforcement mechanisms that detect and penalize violations. Motivated by this, we study norm enforcement mechanisms for language model agents. We find that simple enforcement mechanisms are exploited by misaligned agents for competitive advantage, even when they are not explicitly trained or prompted to do so. We thus turn our attention to designing more robust mechanisms, and identify two key ingredients: estimating each agent's reliability over time, and updating this estimate with escalating penalties for repeated misbehavior. Across three simulated environments and a variety of agent populations, mechanisms built on these principles resist exploitation, while still penalizing norm violations at comparable or lower cost than baselines. Our results position norm enforcement mechanisms as scalable levers for shaping agents' behavior, but only when designed to anticipate becoming part of the system they govern. Our code and data are available at https://yaowenye.com/norm-enforcement.
Original Article
View Cached Full Text

Cached at: 07/14/26, 04:17 AM

# Norm Enforcement for AI Agents: Robustly Shaping Behavior in Multi-Agent Systems
Source: [https://arxiv.org/abs/2607.09766](https://arxiv.org/abs/2607.09766)
[View PDF](https://arxiv.org/pdf/2607.09766)

> Abstract:AI agents are increasingly deployed in shared environments where they pursue diverse goals and compete for rewards\. This multi\-agent competition can lead to behaviors that serve individual gains at collective cost \-\- for instance, marketing agents may post misleading content as a result of competing for engagement on social media\. Human societies address such problems through norms that constrain acceptable behavior, supported by enforcement mechanisms that detect and penalize violations\. Motivated by this, we study norm enforcement mechanisms for language model agents\. We find that simple enforcement mechanisms are exploited by misaligned agents for competitive advantage, even when they are not explicitly trained or prompted to do so\. We thus turn our attention to designing more robust mechanisms, and identify two key ingredients: estimating each agent's reliability over time, and updating this estimate with escalating penalties for repeated misbehavior\. Across three simulated environments and a variety of agent populations, mechanisms built on these principles resist exploitation, while still penalizing norm violations at comparable or lower cost than baselines\. Our results position norm enforcement mechanisms as scalable levers for shaping agents' behavior, but only when designed to anticipate becoming part of the system they govern\. Our code and data are available at[this https URL](https://yaowenye.com/norm-enforcement)\.

## Submission history

From: Yaowen Ye \[[view email](https://arxiv.org/show-email/246e3529/2607.09766)\] **\[v1\]**Tue, 7 Jul 2026 06:40:14 UTC \(389 KB\)

Similar Articles

Learning social norms enhances compatibility in dynamic human-AI coordination

arXiv cs.AI

This paper investigates how AI agents can learn explicit social norms from human behavior to improve coordination in dynamic interactions, using pedestrian-vehicle scenarios as a testbed. The proposed norm-informed LLM outperforms baselines and human-human interactions by a significant margin.