As global energy demand continues to rise and environmental sustainability becomes increasingly critical, optimizing energy use in residential buildings has emerged as a major objective for both governments and the private sector. Traditional energy management strategies often lack the intelligence and adaptability required to meet modern energy efficiency goals. Artificial Intelligence (AI) offers powerful tools for understanding, predicting, and controlling energy consumption in real time. This paper explores the foundational technologies that enable AI-driven energy optimization in smart residential buildings, including machine learning algorithms, Internet of Things (IoT) integration, and predictive analytics. Use cases such as intelligent heating, ventilation, and air conditioning (HVAC) systems, appliance scheduling, and occupant behaviormodeling are examined. Case studies from smart city initiatives and pilot programs illustrate tangible energy savings and improved user comfort. Ethical and technical considerations, such as privacy, data security, and interoperability, are analyzed. Finally, challenges like model generalization, sensor reliability, and realtime responsiveness are discussed alongside future directions, including federated learning, demand-response systems, and personalized energy management. AI-powered energy optimization is central to building the sustainable, adaptive, and efficient residential ecosystems of the future