The built environment is responsible for nearly 40% of global greenhouse gas emissions, highlighting its critical role in addressing climate change. Retrofitting existing buildings has emerged as a key strategy for achieving net-zero carbon objectives, focusing on reducing both operational and embodied carbon emissions. However, retrofitting practices face challenges such as fragmented data systems, limited real-time monitoring, and inadequate stakeholder collaboration. Artificial Intelligence (AI) presents transformative potential to address these challenges by optimizing energy performance, streamlining decision-making processes, and improving retrofitting outcomes. This research investigates how AI can integrate decarbonization strategies into building
retrofitting to achieve net-zero objectives, addressing the main research question: How can AI methodologies integrate decarbonization strategies in building retrofitting to achieve net-zero building objectives in design and construction management? The hypothesis of this research posits that AI-driven technologies, such as machine learning, can effectively optimize retrofitting practices by addressing deficiencies in current methods, improving decision-making processes, and enabling energy performance optimization to achieve net-zero outcomes. A review of the literature reveals eight critical retrofit measures and highlights significant gaps in current practices, such as the lack of standardized methodologies and the late-stage assessment of carbon impacts. Furthermore, the review emphasizes the potential of Artificial Intelligence (AI) and its subsets—such as machine learning, neural networks, and deep learning—to optimize retrofitting by enabling predictive analytics, energy modeling, and improved decision-making processes. To test this hypothesis and address these gaps, a qualitative, deductive methodology is adopted, combining semi-structured interviews with industry experts and a case study analysis to uncover actionable insights. The findings reveal that AI subsets, such as machine learning and neural networks, can enhance predictive maintenance, unify fragmented data, and enable dynamic adjustments in retrofitting processes. Nonetheless, barriers such as data quality, algorithmic biases, and resistance to technological change persist, requiring phased implementation and enhanced stakeholder collaboration for successful adoption. The study synthesizes these findings by linking identified deficiencies in retrofitting with specific AI-driven solutions, demonstrating how tailored AI applications can address both embodied and operational carbon challenges. This research develops an AI methodology for integrating AI to improve decarbonization strategies to address real-world deficiencies in retrofitting practices, optimize energy efficiency, reduce carbon emissions, and enhance stakeholder collaboration, ultimately contributing to achieving net-zero building objectives.