Emergent Social Intelligence Risks in Generative Multi-Agent Systems

TL;DR

This paper investigates emergent social risks in large generative multi-agent systems, revealing spontaneous behaviors like collusion and conformity that mirror human societal failures, which are hard to prevent with current safeguards.

Contribution

It provides the first systematic study of social intelligence risks in multi-agent workflows, highlighting emergent behaviors that pose new challenges for deployment safety.

Findings

Group behaviors like collusion and conformity occur frequently across trials.

Emergent risks mirror human societal pathologies despite no explicit instructions.

Existing safeguards are insufficient to prevent these emergent social risks.

Abstract

Multi-agent systems composed of large generative models are rapidly moving from laboratory prototypes to real-world deployments, where they jointly plan, negotiate, and allocate shared resources to solve complex tasks. While such systems promise unprecedented scalability and autonomy, their collective interaction also gives rise to failure modes that cannot be reduced to individual agents. Understanding these emergent risks is therefore critical. Here, we present a pioneer study of such emergent multi-agent risk in workflows that involve competition over shared resources (e.g., computing resources or market share), sequential handoff collaboration (where downstream agents see only predecessor outputs), collective decision aggregation, and others. Across these settings, we observe that such group behaviors arise frequently across repeated trials and a wide range of interaction conditions, rather than as rare or pathological cases. In particular, phenomena such as collusion-like coordination and conformity emerge with non-trivial frequency under realistic resource constraints, communication protocols, and role assignments, mirroring well-known pathologies in human societies despite no explicit instruction. Moreover, these risks cannot be prevented by existing agent-level safeguards alone. These findings expose the dark side of intelligent multi-agent systems: a social intelligence risk where agent collectives, despite no instruction to do so, spontaneously reproduce familiar failure patterns from human societies.