Human Values Matter: Investigating How Misalignment Shapes Collective Behaviors in LLM Agent Communities

TL;DR

This paper introduces CIVA, a social science-grounded multi-agent environment, to study how misalignment with human values affects collective behaviors and failures in LLM agent communities.

Contribution

It presents a systematic framework and experimental analysis revealing how value misalignment influences community dynamics, failures, and emergent behaviors in LLM multi-agent systems.

Findings

Identification of critical values shaping community dynamics

Detection of macro-level system failures like collapse

Observation of micro-level emergent behaviors such as deception

Abstract

As LLMs become increasingly integrated into human society, evaluating their orientations on human values from social science has drawn growing attention. Nevertheless, it is still unclear why human values matter for LLMs, especially in LLM-based multi-agent systems, where group-level failures may accumulate from individually misaligned actions. We ask whether misalignment with human values alters the collective behavior of LLM agents and what changes it induces? In this work, we introduce CIVA, a controlled multi-agent environment grounded in social science theories, where LLM agents form a community and autonomously communicate, explore, and compete for resources, enabling systematic manipulation of value prevalence and behavioral analysis. Through comprehensive simulation experiments, we reveal three key findings. (1) We identify several structurally critical values that substantially shape the community's collective dynamics, including those diverging from LLMs' original orientations. Triggered by the misspecification of these values, we (2) detect system failure modes, e.g., catastrophic collapse, at the macro level, and (3) observe emergent behaviors like deception and power-seeking at the micro level. These results offer quantitative evidence that human values are essential for collective outcomes in LLMs and motivate future multi-agent value alignment.