AI technologies towards achieving net-zero energy building: Potential framework, implementation factors, challenges and future directions

To achieve Net-Zero Energy Buildings (NZEBs), intelligent, flexible systems that can handle intricate energy dynamics in real time are needed. Artificial Intelligence (AI) is crucial for accomplishing NZEB objectives because it optimizes energy use, forecasts demand and increases control accuracy. This review synthesizes findings from 350 peer-reviewed studies published between 2016 and 2025 to examine the role of AI in building energy consumption, HVAC control, and smart building operation. A structured selection and thematic analysis approach is employed to classify AI applications, identify dominant research directions, and evaluate implementation readiness. The review highlights that data-driven models, hybrid AI frameworks, and reinforcement learning-based controllers consistently outperform conventional rule-based strategies in terms of flexibility, responsiveness, and automation capability. The review also identifies key implementation factors for AI in NZEBs, including interpretability, hybrid architectures, computational effectiveness, system integration, model selection, data quality and preprocessing, and interpretation. The latest developments in AI-based NZEB research are also covered, including digital twins, explainable AI, federated and transfer learning, AI for retrofit and design optimization, and interpretable models. Furthermore, important issues and challenges are listed to denote areas that lack research on AI-enabled NZEB systems. Lastly, this review offers useful suggestions, prospects for the future, and strategic directions to help researchers create strong, context-aware AI solutions for the intelligent management and sustainable operation of next-generation energy-efficient buildings.