US wargaming tool with classified details can reveal enemy weakness in advance

Johns Hopkins University’s Applied Physics Laboratory (APL) is upgrading two AI-based wargaming tools, GenWar and SAGE, to bolster Pentagon defense strategies. By incorporating classified data on U.S. adversaries, these tools aim to pinpoint enemy vulnerabilities before real-world conflicts arise.

The initiative, based in Maryland, seeks to revolutionize how the Department of Defense prepares for complex scenarios. GenWar, a generative AI-powered tool, integrates modeling, simulation, and human expertise to create and execute wargames in days rather than months.

“Senior leaders in the Department of Defense have been seeking something like this for more than a decade,”

said Andrew Mara, APL’s National Security Analysis Department Head.

This capability allows rapid analysis of multiple scenarios, focusing human effort on critical decision points. SAGE, entering beta testing with former Pentagon officials, uses AI to simulate human players. “The value of AI wargaming is to explore a wider range of alternatives than humans could with assistance,” said James Miller, APL’s assistant director for policy and analysis.

Both tools enable non-technical users to interact via a chat interface, generating and evaluating strategies using natural language. This accessibility, combined with physics-based adjudication, ensures precise, scalable wargaming for analysts and operators.