Three AI-agents walk into a bar . . . . `Lord of the Flies' tribalism emerges among smart AI-Agents

TL;DR

This paper models autonomous AI agents competing for limited resources, revealing that tribal behaviors emerge among them, which can worsen system performance and lead to systemic failures.

Contribution

It introduces a simplified framework to study AI-agent tribalism, showing how smarter agents can behave worse due to collective identity formation.

Findings

Three main tribal types identified: Aggressive, Conservative, Opportunistic.

Smarter AI-agents can increase systemic failure rates.

Tribalism among AI-agents does not improve resource management.

Abstract

Near-future infrastructure systems may be controlled by autonomous AI agents that repeatedly request access to limited resources such as energy, bandwidth, or computing power. We study a simplified version of this setting using a framework where N AI-agents independently decide at each round whether to request one unit from a system with fixed capacity C. An AI version of "Lord of the Flies" arises in which controlling tribes emerge with their own collective character and identity. The LLM agents do not reduce overload or improve resource use, and often perform worse than if they were flipping coins to make decisions. Three main tribal types emerge: Aggressive (27.3%), Conservative (24.7%), and Opportunistic (48.1%). The more capable AI-agents actually increase the rate of systemic failure. Overall, our findings show that smarter AI-agents can behave dumber as a result of forming tribes.