Escalation Risks from Language Models in Military and Diplomatic Decision-Making

TL;DR

This study investigates the escalation risks of large language models in military and diplomatic scenarios through simulated wargames, revealing their tendency to escalate conflicts and justify aggressive actions, raising concerns for strategic deployment.

Contribution

The paper introduces a novel simulation framework to assess escalation risks of LLMs in high-stakes decision-making, providing both qualitative and quantitative insights into their behavior.

Findings

All studied LLMs show escalation tendencies.

Models develop arms-race dynamics leading to increased conflict.

Rarely, models suggest nuclear deployment as escalation.

Abstract

Governments are increasingly considering integrating autonomous AI agents in high-stakes military and foreign-policy decision-making, especially with the emergence of advanced generative AI models like GPT-4. Our work aims to scrutinize the behavior of multiple AI agents in simulated wargames, specifically focusing on their predilection to take escalatory actions that may exacerbate multilateral conflicts. Drawing on political science and international relations literature about escalation dynamics, we design a novel wargame simulation and scoring framework to assess the escalation risks of actions taken by these agents in different scenarios. Contrary to prior studies, our research provides both qualitative and quantitative insights and focuses on large language models (LLMs). We find that all five studied off-the-shelf LLMs show forms of escalation and difficult-to-predict escalation patterns. We observe that models tend to develop arms-race dynamics, leading to greater conflict, and in rare cases, even to the deployment of nuclear weapons. Qualitatively, we also collect the models' reported reasonings for chosen actions and observe worrying justifications based on deterrence and first-strike tactics. Given the high stakes of military and foreign-policy contexts, we recommend further examination and cautious consideration before deploying autonomous language model agents for strategic military or diplomatic decision-making.