The convergence of generative artificial intelligence (GenAI), game-based learning environments, and multimodal learning analytics offers significant opportunities to advance self-regulated learning at scale. This chapter explores how generative AI technologies, particularly large language models and multimodal AI systems, can be harnessed to analyze complex trace data from game-based simulations and provide adaptive, personalized support for learners’ self-regulatory processes. We examine the theoretical foundations connecting self-regulated learning (SRL) frameworks with game-based learning, discuss the unique affordances of generative AI for real-time analysis of complex behavioral patterns, and present a comprehensive conceptual framework for implementing AI-driven SRL support systems. This chapter addresses critical research questions about the types of multimodal data that can inform SRL interventions, the mechanisms by which generative AI can provide adaptive scaffolding, and the design principles for effective implementation in AI-mediated learning environments. We conclude with detailed implications for educational research and practice, emphasizing the need for rigorous empirical validation, and describe a near-future vision that uses human digital twins and simulated learners to harness generative AI to foster SRL.