Multidomain Operations Wargame Supported by Generative AI

In September 2025, the Mountain Warfare Centre of Excellence (MW COE) conducted the Mountain Warfare Multidomain Operations (MDO) Wargame in Poljče, Slovenia, in cooperation with the NATO Science and Technology Organization (STO) Systems Analysis and Studies (SAS) 172nd Research Task Group – Multi-Domain Toolkit.

The event aimed to explore how Generative Artificial Intelligence (GAI), including Large Language Models (LLMs) such as BortChat, can support decision-making and enhance understanding of operations across multiple domains in a complex mountain warfare environment.

The MDO Wargame was designed to support the development of the Multidomain Mountain Warfare Capstone Concept and to deepen the understanding of how mountain warfare fits within NATO’s future operational framework.

Aligned with NATO’s vision for Multidomain Operations by 2040, the wargame examined how land, maritime, air, cyber, and space domains can be synchronized in a mountainous battlespace. The experiment focused on applying NATO’s manoeuvre approach in complex terrain and testing AI-assisted decision-making at the tactical level.

The activity was conducted in three main phases:

1. Preparation Phase: Coordination via video-teleconference, introduction by the SAS 172 team, familiarization with NATO ATP 3.2.1.3 Mountain Warfare, and presentation of the experiment’s methodology and objectives.
2. Wargaming Phase: Execution of repeated wargame iterations under varying conditions to test different force compositions and decision-making methods.
3. Analysis and Reporting Phase: Collection and evaluation of experimental data, followed by analysis and report production.

The experiment was based on one historical wargame scenario which was played multiple times under different conditions to evaluate the impact of AI-assisted decision-making and force composition.

• Day 1: The participants played the original historical wargame twice — once without and once with the support of Generative AI (NATO SAS 172 Chatbot). This provided a baseline comparison of manual versus AI-assisted decision-making in a conventional tactical setting.
• Day 2: The same scenario was repeated, but standard infantry units were replaced with mountain infantry units. The objective was to assess how specialized mountain troops, with their unique knowledge, experience, and equipment, would perform in the same operational environment. Again, the scenario was played once without and once with AI support.
• Day 3: The environment was expanded into a MDO setting, aiming to integrate all operational domains — land, air, maritime, cyber, and space — into the wargame. While not all domains could be fully represented in the manual format, the experiment offered valuable insight into the challenges of multidomain integration in a mountainous context.

Each iteration of the wargame was evaluated through key performance indicators: speed of decision-making, quality of decisions, and achievement of victory conditions.

Initial findings suggest that AI-assisted decision-making significantly improved the overall campaign outcomes, especially in complex and dynamic operational environments.
The comparison between traditional and AI-supported approaches showed that GAI can accelerate situational analysis and improve the structure of commander’s intent.

Furthermore, the experiment confirmed that area-specific AI models, tailored to the operational context (in this case, mountain warfare), outperform general-purpose models in providing relevant and tactically sound recommendations.

NATO SAS 172 Chatbot proved to be a valuable decision-support assistant, helping participants synthesize large volumes of data, explore alternative courses of action, and articulate decisions with improved clarity and consistency.

The Mountain Warfare MDO Wargame 2025 successfully demonstrated the potential of GAI as a decision-support tool in NATO’s future operational concepts. The experiment provided essential insights for the ongoing revision of the Mountain Warfare Capstone Concept and underlined the importance of multidomain and multinational collaboration.

The MW COE extends its sincere appreciation to the NATO STO SAS 172 Research Team for their outstanding cooperation and contribution. The Centre looks forward to continuing this innovative partnership and further exploring the role of AI in mountain warfare experimentation and concept development.