Advancing Air Force Wargaming through Artificial Intelligence

A Drive for Transformation

The U.S. Air Force Futures Directorate is actively seeking to revolutionize its wargaming capabilities through advanced artificial intelligence (AI) technologies. This initiative represents a significant shift from conventional, analog methods to a fully digitized, scientific approach aimed at achieving Decision Superiority and integrating an effective Force Design.

In its latest Request for Information (RFI), released on a recent Sunday, Air Force officials emphasized the urgent need for a modernized system that can address current challenges. These challenges include the limitations posed by outdated and vendor-restricted tools, which hinder the ability to conduct coherent analyses of capabilities, Courses of Action (COA), and cost assessments.

Envisioning Revolutionary Solutions

Traditional military wargames often fall short due to their lengthy and expensive nature, as well as their analytical inadequacies. The Air Force now seeks innovative solutions that can facilitate simulations at unprecedented speeds—up to 10,000 times real-time—allowing for rapid scenario evaluations critical to national security strategy.

Desired Technological Capabilities

To realize this ambitious vision, the Air Force has outlined several core requirements for prospective solutions:

• Reinforcement Learning Integration: AI systems should utilize reinforcement learning to model adversary behavior effectively, enabling them to engage with human players and other AI entities.
• Complex Adaptive Simulation: Technologies must represent warfare as a dynamic, complex adaptive system, employing event-driven, agent-based simulations across multiple domains. Each element should act as an independent agent responding in real-time to evolving scenarios.
• Scalability and Security: The envisioned systems should support hundreds of users while accommodating tens of thousands of entities. Moreover, they need to operate across various classification levels, including Unclassified, Secret, and TS/SCI/SAP.

Realism in Simulation

To enhance realism in simulations, the Air Force emphasizes the importance of employing physics-based adjudication mechanisms that accurately depict engagement outcomes and associated effects. Consideration of secondary effects, such as the impact of electronic warfare or cyber incidents on logistics and situational awareness, is also crucial to forming a more comprehensive tactical picture.

Streamlined Decision Support

The Air Force is particularly interested in AI capabilities that can expedite the generation and refinement of COAs. By implementing a neuro-symbolic process, planners could quickly analyze options, assessing associated risks and resource allocations. The ability to pause simulations at key junctures, presenting visualizations of potential outcomes and risk corridors, would facilitate more informed decision-making for commanders.

Integration of Large Language Models

Another aspect of this initiative involves leveraging large language models (LLMs) to produce operational tasking orders in real-time. These models could be used to extract relevant data from selected COAs and align it with joint operational standards. Additionally, LLMs could enable rapid transcription and analysis of qualitative data—such as command discussions—during and following wargames.

A Modular Integration Approach

The Air Force is conceptualizing a “WarEngine” platform aimed at simplifying user interaction with simulation technologies. This modular, cloud-based application intends to serve as an integration hub for digital wargaming. Key features will include:

• Interoperability: The system should integrate seamlessly with existing service-approved tools like AFSIM, NGTS, and JICM, utilizing modular Application Programming Interfaces (APIs) to ensure transparency and replicability.
• User Accessibility: A user-friendly interface will allow military personnel to access and leverage simulation technologies effectively.

Call for Responses

Industry stakeholders are invited to contribute their insights and capabilities in response to the Air Force’s RFI. Responses are expected by January 9, and the Air Force looks forward to collaborating with innovative entities capable of propelling its wargaming to new heights.

In summary, as the U.S. Air Force seeks to modernize its wargaming framework, the integration of sophisticated AI tools holds the potential not only to enhance operational effectiveness but also to reshape the future landscape of military strategy.

Jon Harper is the Editor-in-Chief of DefenseScoop, where he leads a team dedicated to delivering in-depth analysis on military technology and its implications for the defense sector. Follow him on X: @JonHarper.