Strategic Competition in the Age of Artificial Intelligence: Emerging Risks and Opportunities in Military AI Applications

List of Key Chapter Titles

1. Introduction
2. Constructing an Analytical Framework for the Strategic Impact of Military AI
3. Impact at the National Level
4. Impact at the International Level
5. Implications for Different Types of Competition
6. Implications for Different Types of Actors
7. Priority Issues for Attention
8. Lessons Learned from Other Domains
9. Toolkit for Exerting Influence
10. Conclusion and Next Steps

Document Introduction

Artificial intelligence is profoundly transforming all areas of society, economy, and policy, including the domains of defense and security. As a leading proponent of civilian and commercial AI applications, while also committed to the responsible development of defense AI, the United Kingdom urgently needs a clear and in-depth understanding of the emerging risks and opportunities brought by military AI applications, as well as how to optimize related response strategies through international cooperation.

In March 2024, the UK Ministry of Defence's Defence AI and Autonomy Unit (DAU) and the Foreign, Commonwealth & Development Office (FCDO) jointly commissioned RAND Europe to conduct this exploratory study. It aims to break through the current research limitations that often focus on tactical levels or non-military issues, analyzing the potential impact of military AI from a strategic height and constructing a corresponding conceptual framework. The study was completed by a transatlantic RAND team, drawing on RAND's 75 years of experience providing research support to multiple governments, combined with its deep accumulation since the 1950s in areas such as the military-strategic impact of AI, game theory, and deterrence theory.

The research adopted a multi-method approach, including a narrative review of approximately 200 core documents (from an initial longlist of 1500 documents), semi-structured interviews with over 50 stakeholders and experts from government, the UN, NATO, the defense industry, AI companies, academia, and other fields, as well as the integration of insights from 7 external workshops, webinars, and 2 parliamentary inquiries. The conceptual framework was iteratively refined through consultations with the UK MoD and FCDO.

The conceptual framework proposed in the report starts from the net effect at the global level, focusing on core dimensions such as the international system, strategic competition dynamics, and the advantage potential and conversion capabilities of various actors. It covers multiple impact areas including society, economy, defense enterprise, and military capabilities, as well as the full continuum of cooperation, competition, and conflict. The study identifies key issues such as information manipulation, empowerment of non-state actors, shifts in offense-defense balance, impact on nuclear deterrence stability, and potential risks of Artificial General Intelligence (AGI). It analyzes the specific impacts of military AI on national strategic advantage, changes in the international system, and various competitive scenarios, while also exploring the differentiated situations faced by countries of different sizes, different governance models, and non-state actors.

Based on risk management experiences from other domains, the report proposes a toolkit containing three types of mechanisms: mechanisms to enhance the application and benefits of defense AI, mechanisms to limit adversaries (including non-state actors and hostile states) from gaining related advantages, and mechanisms to shape global governance arrangements for military AI. These findings provide important theoretical support and practical reference for the UK and other countries in formulating relevant international engagement strategies, optimizing defense AI development pathways, and managing potential risks. Follow-up research will further deepen the detailed exploration of each issue.