Claude Goes to War: The “AI Arms Race” and Frontier AI Companies as New Agents of War

2026 Jeffrey Grey Oration
Claude Goes to War: The “AI Arms Race” and Frontier AI Companies as New Agents of War

In February 2026, Claude went to war. The United States’ reliance on the AI model to support decision making in the war with Iran represents a watershed moment in terms of how states initiate and conduct war. It should also profoundly inform how we understand the emerging world order and which actors will contribute to creating it.

In the 2026 Grey Oration, Professor Toni Erskine will argue that the frontier AI companies that provide such models – including Anthropic, OpenAI, and Google DeepMind – are poised to become new agents of war. In defending this position, she analyses four inter-related phenomena involving these companies: 1) their stark policy reversal in allowing military applications of their AI systems and how they justify this change; 2) their contribution to the accelerating and increasingly militarized “race” over AI innovation; 3) the state’s growing, if reluctant, dependency on these companies; and, perhaps most significantly, 4) their on-going, often invisible, control over the models that the state relies on in war and resulting influence over military judgment.

Together these phenomena challenge the state’s claim to exclusive discretion over its use of force in a way that demands attention. They also raise complex questions of where responsibility lies for reining in the use of these models when the perceived AI “race” prompts military applications for which the technology is not yet ready – or credibly poses catastrophic risks.
 

Speaker

Toni Erskine is Professor of International Politics at the Australian National University (ANU), the 2026 Jeffrey Grey Distinguished Chair at the Australian Defence College, and Associate Fellow of the Leverhulme Centre for the Future of Intelligence at Cambridge University. She is Chief Investigator of the'Anticipating the Future of War: AI, Automated Systems, and Resort-to-Force Decision Making' Research Project,funded by the Australian Government through a grant from Defence, and recipient of the International Studies Association’s 2024-25 Distinguished Scholar Award in International Ethics. In 2025 she was elected Fellow of the Australian Institute of International Affairs (AIIA).She recently served a five-year term as Director of the Coral Bell School of Asia Pacific Affairs at the ANU. She also served as Editor of International Theory: A Journal of International Politics, Law, and Philosophy and Academic Lead for the United Nations (UN) Economic and Social Commission for Asia and the Pacific / Association for Pacific Rim Universities “AI for Social Good: Strengthening Capabilities and Government Frameworks in Asia and the Pacific” Research Project.

Professor Erskine’s research is located at the intersection of international relations, international security, and moral and political philosophy. Her current research addresses the impact of artificial intelligence (AI) on war, including: the use of AI-enabled decision-support systems for resort-to-force decision making; human-machine interaction and its implications for the exercise of restraint; the problem of “misplaced responsibility” to machines; the effects of AI systems on the organisational and deliberative structures of the state; and the roles and responsibilities of frontier AI companies as emerging “agents of war”.