Artificial Intelligence in ESG Reporting: A Scopus-Based Bibliometric Analysis and Conceptual Model for Data-Driven Decision Support

Abstract

The purpose of this manuscript is to review the role of AI in evolving ESG reporting from a historical, compliance-driven approach toward an integrated analysis and decision-support system. The analysis is based on a bibliometric analysis of manuscripts gathered from the Scopus database, with 765 publications from 2004–2026. It was performed keyword co-occurrence analysis using VOSviewer to reveal the main thematic clusters and structure of the research field. An analysis was also performed by using the Ishikawa diagram and a Pareto–Lorenz analysis to determine whether this degree of concentration was indicative of the relative importance among key concepts. The results reflect an increased role for technologies like machine learning and natural language processing in the analysis of ESG data, notably unstructured information. Among the most common were key drivers of AI adoption, such as regulatory pressure, stakeholder expectations and the desire to improve data quality. In this context, the manuscript seeks to propose a conceptual model for AI–driven transformation of ESG reporting, whereby AI serves as an integrating factor connecting data processing and analytics, decision-making and due diligence and enforcement in practice.