TLDR: The Johns Hopkins University Applied Physics Laboratory (JHUAPL) is advancing its AI-enhanced wargaming tools, GenWar and SAGE, for deployment on classified networks within the Department of Defense, Energy Department, and the Intelligence Community. These systems aim to leverage generative AI for strategic simulations using highly sensitive data, while mitigating risks associated with AI ‘hallucinations’.
The Johns Hopkins University Applied Physics Laboratory (JHUAPL) is at the forefront of developing advanced artificial intelligence-driven wargaming tools, GenWar and the Strategic AI Gaming Engine (SAGE), for deployment on classified networks of the U.S. Department of Defense (DoD), Energy Department (DoE), and the Intelligence Community (IC). This initiative seeks to harness generative AI to simulate complex strategic scenarios using highly classified data concerning adversaries, while simultaneously safeguarding against the AI’s tendency to generate inaccurate or nonsensical information.
According to James Miller, JHUAPL’s assistant director for policy and analysis, these AI-enhanced wargaming tools are being upgraded to operate on Top Secret/Sensitive Compartmented Information (TS/SCI) networks. Miller stated at a recent wargaming conference, “It’s going to be somewhat straightforward to move this onto TS/SCI networks. We have one sponsor that’s very keen to do it quickly and others are interested as well.” He emphasized the critical role of wargaming in strategic planning, from long-term budget assessments to real-time crisis response.
GenWar, one of the two primary systems, is designed to streamline the wargame setup process. It integrates large language models (LLMs) with high-fidelity simulation tools, such as the Advanced Framework for Simulation, Integration, and Modeling (AFSIM). This allows users, including policymakers and commanders, to articulate desired simulations in plain English, which the AI then translates into actionable inputs for the simulation in minutes, a process that traditionally took months. Andrew Mara, head of national security analysis at JHUAPL, highlighted GenWar’s built-in safeguards, noting, “It forces the AI to only do things that follow the laws of physics. It can’t spin off into ‘I landed 16 aircraft on the moon.’”
The other tool, SAGE, takes an even more ambitious approach by using generative AI to replace human players in wargames. It can simulate high-level policy discussions, such as National Security Council meetings, where AI-driven chatbots assume the roles of various policymakers, arguing out courses of action. This capability aims to provide deeper insights by analyzing and explaining AI-generated decisions.
Also Read:
- U.S. Army Accelerates Autonomous Drone Deployment with New Solicitation for ‘Launched Effects’ Systems
- Agentic AI’s Dual Edge: Accelerating Chip Design Amidst Security Concerns
The integration of AI into wargaming significantly reduces the time required for planning and executing these exercises, from several months to as little as a few days or weeks. This efficiency gain is crucial for accelerating mission analysis and providing decision-makers with rapid insights. The long-term vision for these capabilities is to create tools that allow decision-makers at various levels to run numerous wargame repetitions on personal computers, making strategic insights more accessible and timely. This development aligns with broader efforts by the Trump Administration to accelerate American AI capabilities, particularly in areas with high payoff and manageable risks like wargaming.
