Artificial Intelligence Applications in Energy Systems: A Comprehensive Review

The global energy sector is transforming under decar-bonization, decentralization, and digitalization, introducing complex-ity and volatility into system operations, challenging the limits of tra-ditional analytical methods. Artificial intelligence (AI) encompassing machine learning, deep learning, and reinforcement learning has emerged as a critical enabler for building the next generation of energy infrastructure. This review provides a systematic assessment of AI's role across key domains including forecasting, optimization, con-trol, and maintenance. It also examines the emerging potential of large language models from several application perspectives. Our analysis identifies a clear trend from data-driven models toward more sophisticated hybrid and physics-informed architectures. We docu-ment this progression through specific advances in forecasting elec-tricity load, renewable generation, and electricity price, in optimizing grid scheduling, energy storage and demand-side resources, and in enhancing system resilience through intelligent control and mainte-nance. Beyond current applications, we discuss future research chal-lenges: developing trustworthy AI for high-stakes environments, cre-ating scientific AI that integrates physical principles, managing com-plex sector-coupled systems, and progressing toward generalized energy foundation models. Overall, this review offers a structured understanding of the synergy between AI and energy systems, outlin-ing current capabilities and the governance required for trustworthy and equitable deployment.