IA generativa em finanças: uma análise bibliométrica com VOSviewer

Abstract

Generative Artificial Intelligence (GenAI) has been transforming the financial sector by expanding the possibilities of automation, data analysis, and decision-making support, driving the growth of scientific production on the subject. This study aimed to analyze the evolution of the literature on GenAI applied to finance through a bibliometric analysis combined with content analysis. The research adopted the PRISMA protocol for study selection, using the Scopus database and VOSviewer software. Initially, 157 records were identified, of which 41 articles published between 2023 and April 2026 comprised the final corpus of the analysis. The results showed the rapid growth of publications, with a predominance of studies originating from India, the United States, and the United Kingdom, in addition to the consolidation of the literature into four thematic clusters: applications of generative AI in the financial sector; modeling, security, and governance of intelligent systems; large language models and natural language processing; and ethical, regulatory, and socioeconomic aspects of generative AI. An evolution in scientific production was observed, initially focusing on technological foundations and shifting toward research on organizational applications, governance, security, regulatory compliance, and the responsible use of the technology. The analysis also revealed gaps related to the long-term impacts of GenAI adoption, human-AI collaboration, the governance of intelligent systems, and the empirical evaluation of security mechanisms in financial environments. As a contribution, this study systematizes scientific knowledge on Generative Artificial Intelligence applied to finance, identifies emerging trends, highlights research gaps, and proposes a research agenda capable of guiding future studies and supporting researchers, organizations, and policymakers in the responsible development and adoption of this technology.