DEA Meets AI in the Context of M&A: A Three-Decade Evolution and Research Outlook

The In the past three decades (1998–2025), the intersection of Data Envelopment Analysis (DEA), Artificial Intelligence (AI), and Mergers and Acquisitions (M&A) has evolved into a specialised yet expanding research domain. This study conducts an intensive bibliometric and thematic analysis of 480 scientific publications indexed in the Web of Science, spanning 259 sources and exhibiting an annual growth rate of about 14 percent. Using Biblioshiny, we map the evolution of intellectual contributions, methodological trends, and collaborative networks, including patterns of international co authorship. The results reveal three distinct phases. The first is dominated by conventional DEA models applied mainly to post merger efficiency evaluation. The second is marked by sectoral specialisation, particularly in banking. The most recent phase reflects early but growing attempts to embed AI techniques, such as machine learning and random forests, into DEA based frameworks. Despite this expansion, DEA-AI integration remains limited, and applications continue to focus predominantly on post-merger assessment, even with the growth of scholarly production and global engagement, especially from Asia and Latin America. Moreover, the ESG (Environmental, Social, and Governance) dimension is largely absent, exposing a critical gap at the convergence of performance efficiency and sustainable strategy planning. This study consolidates the current state of the field and proposes a forward looking research agenda that prioritises hybrid DEA-AI methodologies, pre-merger assessments, and ESG inclusive models in order to support more intelligent, accountable, and future oriented M&A decisions.