Exia Labs Announces Cooperative R&D with U.S. Military Academy to Advance AI Red Agents for Strategic Wargaming
The team will develop AI agents to simulate adversary strategic doctrine and decision-making at scale
BELLEVUE, Wash., Feb. 5, 2026 – Exia Labs, a leader in spatial intelligence and autonomous agent development for mission-critical environments, today announced a Cooperative Research and Development Agreement with the U.S. Military Academy at West Point. The collaboration will focus on developing AI agents to simulate near-peer adversaries in contested multi-domain environments.
Under the agreement, Exia Labs will work closely with the West Point Simulation Center and the Defense and Strategic Studies Program. Together, the teams will develop Red agents capable of navigating wargame environments while modeling the behaviors of strategic competitors.
This partnership aligns with the Department of War’s designation of Applied Artificial Intelligence (AAI) as a Critical Technology Area to achieve decision superiority. The initiative reflects the growing need for computational software agents that emulate adversarial human decision-making. The U.S. Army’s Red Team Handbook notes that human Red players face limitations in maintaining adversary perspectives over extended periods due to cultural biases, fatigue, and unconscious adoption of friendly force thinking.
“By deploying AI agents that function as Red commanders, we are moving beyond the ‘Western-centric lens’ that often flaws our understanding of adversarial motivations,” said Lt. Col. Jessica C. Caddell, Director Defense and Strategic Studies. “These agents allow cadets to conduct virtual consultations with a thinking enemy, receiving immediate responses rooted in genuine military reasoning and historical doctrine rather than our own friendly-force assumptions.”
AI-native Red agents provide a critical solution to the “human labor bottleneck” by automating the role of the adversary with high-fidelity doctrinal precision. Unlike human role-players who are susceptible to cognitive fatigue and “mirror-imaging,” AI agents remain doctrinally anchored and unbiased over extended periods. These agents allow for statistical wargaming, enabling warfighters to pressure-test thousands of “what-if” scenarios against a pacing challenge like the PRC, which would be impossible to execute manually. These agents will be used to train the next generation of military leaders and stress-test strategic assumptions.
“By integrating Exia Labs’ Red Agents with the world-class expertise of West Point’s Defense and Strategic Studies program and their Simulation Center, we are creating a unique, high-fidelity feedback loop that is essential for tuning adversarial AI,” said Jonathan Pan, CEO at Exia Labs. “The Academy’s deep understanding of peer-competitor doctrine allows us to refine our agents’ decision-making logic with unprecedented precision. This collaboration doesn’t just provide a tool for wargaming; it ensures that the next generation of Army leaders is training against a digitally replicated adversary that thinks, adapts, and competes with the same strategic complexity they will face in the field.”




Looking foward for the opening of the applications.