How NATO can integrate AI to prevail in future algorithmic warfare

(NATO)

By Dominika Kunertova • Mar 31, 2026

Military artificial intelligence (AI) is moving from the margins of experimentation into the core of how NATO will fight, make critical decisions and deter competitors over the next decade. The 2022 NATO Strategic Concept identifies the technological edge to be critical for the Alliance to fulfil its core tasks. Both contemporary warfare and renewed strategic competition suggest that data-driven AI decision-support systems and autonomous battlefield capabilities augmented with AI will define the character of future conflicts. There is a justified focus on evaluating strategic risks associated with such systems.

This report argues that integrating AI into military systems does not generate vulnerabilities that are fundamentally new in kind compared to existing cyber risks. But the difference lies in consequences. Once AI-enabled decision-support systems and autonomous platforms become critical to Alliance operations, interference with data, models, and computing infrastructure may have implications for NATO’s ability to see, decide and act under pressure. Similarly, the offensive use of AI-enabled capabilities does not, on its own, raise or lower the nuclear threshold. Escalation thresholds in algorithmic warfare will continue to be driven by effects on the ground rather than by whether a system is AI-enabled. Yet the characteristics of AI—the speed, system opacity and physical infrastructure—create more room for human error, misperception, and miscalculation.

To explore such possibilities, the Atlantic Council’s Transatlantic Security Initiative, in partnership with the NATO Office of the Chief Scientist, conducted a foresight study to clarify how adversaries might counter AI-enabled capabilities and to examine what this means for NATO doctrine, strategy and deterrence. The research combined horizon scanning and expert interviews, an off-the-record workshop held in Washington, under Chatham House rules, and scenario modeling. The project mapped AI technology trends across decision-support systems and autonomous platforms, identified likely AI vulnerabilities and vectors of attack, and explored escalation dynamics through structured discussion and scenario-based exercises.