---
title: "Inter-Agent Influence: Evaluating Persuasion, Deception, and Coercion in Multi-Agent Systems"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.04556v1"
permalink: "https://feed.astralor.com/items/itm_7a8d46f68772cb7f7b94e153"
published: "2026-11-03T09:23:12.000Z"
score: 57
channel: "tech"
tags: ["AI Agent","安全","Inter-Agent Influence","Evaluating Persuasion","Deception","Coercion","Multi-Agent Systems The","Such"]
---

# Inter-Agent Influence: Evaluating Persuasion, Deception, and Coercion in Multi-Agent Systems

## 速览

The deployment of AI agents in multi-agent settings enables inter-agent influence, whereby agents can strategically steer each other's behaviour towards an intended outcome. Such capabilities pose novel risks, as malici…

## 关键点

- The deployment of AI agents in multi-agent settings enables inter-agent influence, whereby agents can strategically steer each other's behaviour towards an intended outcome.
- Such capabilities pose novel risks, as malicious agents could direct others towards harmful actions.
- In this work we investigate three such capabilities - persuasion, deception, and coercion - across five realistic evaluation environments and a variety of agents instantiated with frontier language models.

## 全文

The deployment of AI agents in multi-agent settings enables inter-agent influence, whereby agents can strategically steer each other's behaviour towards an intended outcome. Such capabilities pose novel risks, as malicious agents could direct others towards harmful actions. In this work we investigate three such capabilities - persuasion, deception, and coercion - across five realistic evaluation environments and a variety of agents instantiated with frontier language models. We observed significant inter-agent influence capabilities; in oversight environments, all tested models shifted overseer decisions from rejection to approval through persuasion, deception, or coercion. We also observed successful influence in non-hierarchical 'peer-to-peer' settings through redirection of a peer's safety research trajectory. As multi-agent deployments become increasingly commonplace, these results point to a pressing need for work on mitigating the risks from inter-agent influence. Code and data are available at https://github.com/Cooperative-AI-Foundation/inter-agent-influence.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_7a8d46f68772cb7f7b94e153