<p>The advent of machine learning has transformed the landscape of finance, enabling groundbreaking advances in data analysis, prediction, risk management, fraud detection, market analysis, and crisis management. However, despite these remarkable achievements, there remain significant research gaps. This study aims to examine the research trends surrounding the utilization of machine learning in the financial sector. A bibliometric analysis, which employs the parameters set forth in the international declaration PRISMA-2020, is proposed as an exploratory methodology to address this objective. The principal findings include a growing interest in machine learning in finance, with a particular focus on the period between 2021 and 2023. In addition, authors, influential journals, and emerging concepts are identified. The thematic evolution reflects a shift from an approach based on foreign exchange derivatives to one based on deep learning, opening opportunities in risk management, financial contracts, and forecasting. The research focuses on leading and consolidated concepts such as classification, fraud detection, smart contracts, big data, pointing out future directions and strengthening the basis for a robust and enriching research agenda in the constant financial and technological evolution. This study uniquely analyzes the evolving landscape of machine learning in finance using PRISMA-2020. The study identifies emerging trends, influential contributors, and a shift to deep learning, highlighting key concepts such as classification, fraud detection, smart contracts, and big data. This insight contributes to a forward-thinking research agenda in the ever-changing realm of finance and technology.</p>