Journal of Research in Information Technology and Its Applications ISSN: 3139-504X (Online)

Authors: Gajender Sharma, Manish Verma, Pooja Kulkarni, Arvind Chauhan

Abstract: Financial fraud has become a major challenge for banking and digital financial ecosystems due to rapid growth of online transactions, e-commerce, and mobile payments. Traditional rule-based fraud detection systems often fail to detect complex and evolving fraud patterns. Artificial Intelligence (AI) techniques such as machine learning, deep learning, and anomaly detection have emerged as powerful tools to enhance fraud detection accuracy and efficiency. This paper reviews the concept, techniques, applications, and challenges of AI-based fraud detection in financial systems. It explains how AI models analyse large transaction datasets, detect suspicious patterns, and support real-time monitoring. The study also highlights implementation issues such as data quality, model interpretability, privacy concerns, and adversarial fraud tactics. The paper concludes that AI-driven fraud detection significantly improves financial security and risk management, though integration with regulatory and ethical frameworks remains necessary.

Keywords: Artificial Intelligence, Fraud Detection, Financial Systems, Machine Learning, Anomaly Detection, Financial Technology