Critical review of using artificial intelligence for environmental, social and governance (ESG) assessment in construction

The rise in the use of artificial intelligence (AI) in every facet of life is associated with some benefits and constraints, especially in non-information technology-related fields. This study evaluates the current landscape of artificial intelligence (AI) techniques, drivers and barriers to environmental, social and governance (ESG) assessments within construction project management (PM). The study adopted a systematic literature review approach. It involved formulating research questions, identifying search keywords, collecting and screening the articles and thoroughly examining the literature to determine the emerging AI techniques, critical drivers and barriers to AI adoption for ESG assessment and proposing future research avenues. Using Scopus, 129 related papers were selected for the review. ESG assessment in construction PM has significantly relied on machine learning techniques, with environmental and governance areas being the primary focus. Some emerging techniques include deep learning for image recognition and natural language processing for text analysis. The top three drivers for adopting AI for ESG assessment are the reduction in human errors in ESG assessment, improved efficiency in ESG assessment and efficient data management. The top three barriers are a lack of critical ESG data/information, inadequate AI construction expertise and the high cost of maintaining AI systems. The findings of this research will inform project professionals and policymakers about the innovative and reliable AI technologies they can utilise in projects to achieve more sustainable outcomes. Overall, the study presents a collection of observations, propositions and pathways for future efforts that examine the contribution of AI-driven PM research to ESG issues.