As defense budgets are cut, assets and personnel are being increasingly stretched to meet operational tasks, making it ever more difficult to allocate platforms and subject matter experts (SMEs) to train the next generation of operators. One key area where there is a shortage of platforms and experienced SMEs is in anti-submarine warfare (ASW). At the very time when many countries are purchasing sophisticated submarines and the potential submarine threat is increasing, fewer operational submarines and SMEs are available for training tasks. An innovative solution to these shortages is to use an ‘expert system’ ASW simulator employing automated intelligent entities to generate realistic threat actions. This innovative solution provides added benefits as it improves the quality of simulator training whilst reducing the workload on the available ASW simulator instructors. This Paper describes the process used for collecting expert knowledge and then using that knowledge to create the automated intelligent behaviors employed by automated intelligent entities in simulators. The collection process enables SMEs to ensure the behaviors represent tactically realistic actions and, for the highest quality simulation, that they do not become predictable. Therefore, for any tactical situation, the system must select the most appropriate behavior and the entity should react realistically to the tactics employed against it by the student. Such autonomous entities allow instructors to perform complex maneuvers and actions with a low level of interaction with the simulation. An additional benefit of the low level of interaction with the simulation is the reduction in the instructor's workload, giving them more time to focus on the overall simulator exercise objectives. As an illustrative example, we present a case study of a system created for the Royal Norwegian Navy (RNoN), which now uses such automated intelligent behaviors in its ASW simulator.