Traditional modeling of civilians walking down a road, ignoring traffic, gunshots, and explosions is worthless, even at scale. Training our warfighters for today must include more than “Move, Shoot, Communicate.” Warfighters need simulations that model human behavior and reactions with accuracy, and this accuracy relies on delivering immense scale. Simulations capable of complex, reactive behaviors of 5M+ civilians civilians in a city are necessary to prepare our warfighters for the challenges they face today, and in the future. While artificial intelligence (AI) is capable of producing assets on a massive scale, incorporating these assets into extended reality (XR) simulations introduces a new set of challenges. Designers and engineers of these simulation systems need to be able to engage and influence “populations” of AI-generated entities into a virtual training environment, which requires interoperability with other complex M&S technologies. This TalX will explore how to incorporate human needs, goals, fears, and sentiment into simulation technologies and provide an example delivered to the UK MoD and Western Allied Nations.