Human–AI Collaboration in Financial Decision-Making: A Comprehensive Bibliometric Analysis (2015–2025)

Abstract

Human-AI collaboration has led to transformative innovations in financial decision-making in areas like investment analysis, fraud detection, financial literacy, cognitive assessment and behavioural finance, and remarkable growth has been witnessed. However, there is insufficient evidence of an accompanying robust bibliometric record unifying co-authorship patterns, country collaborations, evolutions of keywords, and changes of fields. The methodology taken by the study is bibliometric, based on data sourced from the Dimensions database, from 2015 to 2025, and by utilising visualisation tools, to delineate intellectual domains and forge new thematic groups, and facilitate academic collaborations, such as VOSviewer software. The study focuses on 7,742 articles in the commerce and management disciplines, with an additional 5,165 for information systems research, and 2681 for economics literature. The United States, China and Europe in general present evidence of strong governance of published articles and collaborative relationships. Furthermore, keywords also indicate clusters of convergences of cognitive decline with manipulation through technology in economics and progress of AI-based applications, as enhanced diagnostic accuracy in the analysis of financial strategies, paving pathways for interdisciplinary collaborations and to guide cross-disciplinary expertise. Research indicates opportunities such as continued investigation of the ethical implications of artificial intelligence in finance, additional development of models of hybrid intelligence that incorporate multiple forms of cognition in decision-making, and ways to translate complex data in finance so artificial intelligence can make sense of it. The paper proposes possibilities for future research related to the broad area of ethical AI in finance, hybrid intelligence models, and AI explainability in finance.