The U.S. Air Force has conducted a military exercise simulating a high-pressure combat environment, in which artificial intelligence (AI) was used to support the targeting process. The four-day simulation, dubbed “Experiment 3,” was described as the most innovative planning and execution method ever implemented by the Air Force.
Organized by the 805th Combat Training Squadron—known as the Shadow Operations Center at Nellis Air Force Base—participants used AI-powered software to accelerate decision-making and targeting processes.
Lt. Col. Shawn Finney, commander of the squadron, stated: “We’re not just testing software here; we’re questioning assumptions, validating tactics, and shaping the operational architecture that we and our allies will rely on in future conflicts. This was a proving ground for the kill chain of tomorrow.”
One of the main goals of the exercise was to speed up military decision-making while reducing the cognitive load on personnel. AI systems supported the process by providing real-time recommendations to dynamic targeting teams. The software used was developed under the U.S. Department of Defense’s Maven Smart System program, which analyzes incoming data to prioritize targets.
During the exercise, AI also played an active role in creating a command-and-control chain, as well as in tasks related to intelligence, surveillance, reconnaissance, and targeting. The ultimate aim was to build a resilient, data-driven, and automated kill chain at both operational and tactical levels.
Observer teams analyzed the differences between the decisions made by experienced personnel and those suggested by AI. It was emphasized that human intuition, experience, and situational awareness served as a complement to the AI-generated outputs. More accurate assessments from human operators were used to improve the AI algorithms.
Lt. Col. Finney concluded, “This battle lab enables us to test tools before fielding them. This experiment brought us one step closer to the future of command and control.”
