Artificial Intelligence-Based Forecasting of the Impact of Clean Energy Adoption on Property Values: A Systematic Review of Global Evidence

The accelerating global transition towards low-carbon economies has fundamentally reshaped the role of the built environment in achieving climate mitigation objectives. Residential and commercial buildings account for a significant proportion of global energy consumption and greenhouse gas emissions, positioning clean energy adoption within property markets as both an environmental imperative and an economic concern. In recent years, governments, financial institutions, and market participants have increasingly linked energy efficiency improvements, renewable energy installations, and smart energy technologies to property value formation. However, empirical evidence on the extent, persistence, and distribution of these valuation effects remains fragmented and methodologically inconsistent. Traditional econometric approaches, particularly hedonic pricing models, have dominated this research domain but often fail to capture the non-linear, spatially heterogeneous, and temporally dynamic nature of property markets. In response, a growing body of literature has introduced artificial intelligence (AI) and machine learning techniques to enhance forecasting accuracy and interpret complex interactions between energy attributes and property prices.
This systematic review synthesises global empirical and methodological evidence on the application of AI-based models to forecast the impact of clean energy adoption on property values. Drawing on peer-reviewed studies from real estate economics, energy policy, urban analytics, and data science, the review critically evaluates theoretical foundations, modelling frameworks, data sources, and empirical findings. The analysis reveals consistent evidence of positive value capitalisation associated with energy efficiency and renewable energy adoption, commonly described as a “green premium,” although its magnitude varies considerably across regions, market segments, and regulatory contexts. AI-based models, particularly ensemble learning methods and neural networks are shown to outperform traditional econometric approaches in predictive accuracy and their ability to capture non-linear effects. Nevertheless, significant challenges persist, including limited causal inference, data quality constraints, algorithmic bias, and insufficient integration of ethical and equity considerations.
The review concludes that artificial intelligence has become an indispensable analytical tool for understanding sustainable property valuation, but its application must be accompanied by transparent, explainable, and policy-aware frameworks. Future research directions are proposed to advance causal modelling, long-term forecasting, and responsible AI deployment in real estate markets undergoing energy transition.