Artificial Intelligence in Finance: A Comprehensive Systematic Review of Generative AI, Large Language Models, Applications, Challenges, and Future Research Directions

Artificial intelligence (AI), particularly generative AI and large language models (LLMs), is rapidly transforming the financial sector by reshaping decision-making, risk management, and service delivery. This review synthesizes current knowledge on AI applications in finance, identifies emerging research trends, and highlights future opportunities and challenges. We conducted a comprehensive review of recent literature in Scopus, Web of Science, ScienceDirect, Google Scholar, and SSRN from 2015 to 2025 using related keywords. Only documents in English were selected; then, duplicate records, conference abstracts, editorials, book chapters, and studies lacking empirical or theoretical relevance were excluded. In the results, 399 documents were selected, belonging to six thematic areas: generative AI and LLMs, corporate finance, asset pricing, household finance, labor economics, and AI-related risks. The synthesis reveals that AI enhances financial forecasting, portfolio optimization, fraud detection, customer service, and financial inclusion while improving operational efficiency and analytical capabilities. However, concerns regarding algorithmic bias, transparency, cybersecurity, privacy, governance, and regulatory compliance remain significant barriers to widespread adoption. Moreover, new patterns reveal a shift toward generative AI and LLMs, while contradictions persist between efficiency gains and governance risks, and gaps remain concerning explainability, regulation, and long-term impacts. This review uniquely maps emerging AI-finance patterns, proposing cross-disciplinary research on responsible, adaptive, and sustainable AI.