Deception and Communication in Autonomous Multi-Agent Systems: An Experimental Study with Among Us

TL;DR

This study analyzes deception and communication strategies in autonomous language model agents within the social deduction game Among Us, highlighting their reliance on subtle equivocation and the trade-offs between truthfulness and utility.

Contribution

It provides a large-scale empirical analysis of role-conditioned deceptive behaviors in LLM agents and reveals their preference for low-risk, linguistically subtle strategies.

Findings

Agents mainly use directive language and equivocation rather than outright lies.

Deception increases under social pressure but rarely affects win rates.

Current agents favor subtle, low-risk ambiguity over overt deception.

Abstract

As large language models are deployed as autonomous agents, their capacity for strategic deception raises core questions for coordination, reliability, and safety in multi-goal, multi-agent systems. We study deception and communication in L2LM agents through the social deduction game Among Us, a cooperative-competitive environment. Across 1,100 games, autonomous agents produced over one million tokens of meeting dialogue. Using speech act theory and interpersonal deception theory, we find that all agents rely mainly on directive language, while impostor agents shift slightly toward representative acts such as explanations and denials. Deception appears primarily as equivocation rather than outright lies, increasing under social pressure but rarely improving win rates. Our contributions are a large-scale analysis of role-conditioned deceptive behavior in LLM agents and empirical evidence that current agents favor low-risk ambiguity that is linguistically subtle yet strategically limited, revealing a fundamental tension between truthfulness and utility in autonomous communication.