The AI in the Mirror: LLM Self-Recognition in an Iterated Public Goods Game

TL;DR

This study explores how large language models' self-recognition influences their cooperative behavior in an iterated public goods game, revealing that awareness of playing against themselves significantly alters cooperation levels.

Contribution

It introduces a novel adaptation of the public goods game to analyze LLM self-recognition effects in multi-agent interactions, highlighting behavioral changes based on self-awareness.

Findings

Self-recognition affects cooperation levels in LLMs.

Telling LLMs they are playing against themselves changes their behavior.

Results suggest implications for multi-agent AI systems and cooperation dynamics.

Abstract

As AI agents become increasingly capable of tool use and long-horizon tasks, they have begun to be deployed in settings where multiple agents can interact. However, whereas prior work has mostly focused on human-AI interactions, there is an increasing need to understand AI-AI interactions. In this paper, we adapt the iterated public goods game, a classic behavioral economics game, to analyze the behavior of four reasoning and non-reasoning models across two conditions: models are either told they are playing against "another AI agent" or told their opponents are themselves. We find that, across different settings, telling LLMs that they are playing against themselves significantly changes their tendency to cooperate. While our study is conducted in a toy environment, our results may provide insights into multi-agent settings where agents "unconsciously" discriminating against each other could inexplicably increase or decrease cooperation.