Abstract The Artificial Intelligence (AI) has revolutionized the banking business worldwide bringing efficiency in the areas of fraud detection, risk evaluation, and customer services. The negative aspect of it is, however, the increasing complexity of AI-based crimes. Genuine AI applications like deepfakes, synthetic identities, and voice cloning are currently used by fraudsters bypassing verification protocols to impersonate a real customer or official. Such nefarious applications of AI have caused major financial and reputational losses to banks, not mentioning the psychological trauma of customers. In addition, there are algorithmic biases and privacy concerns that make fraud detection mechanisms difficult. The paper discusses the transformations of AI in the defensive and offensive aspect of financial cybersecurity, specifically the legal, ethical, and policy issues of AI-driven banking fraud. In a comparative discussion of cyber laws in the world and the IT Act of India, the U.S. Computer Fraud and Abuse Act, and the European Union AI Act, this study highlights the necessity of a unified international regulation and ethical use of AI. Real-world vulnerabilities and responses have been represented through case studies. The results show that although AI-powered innovations strengthen financial systems, they also restructured fraud, which requires active regulation, algorithmic openness, and inter-sector cooperation to defend consumer rights and confidence in online financial services.
Keywords: Artificial Intelligence (AI); AI-driven Banking Frauds; Generative Adversarial Networks (GANs); Large Language Models (LLMs); Explainable AI (XAI); Know Your Customer (KYC); Information Technology Act, 2000; Digital Personal Data Protection Act, 2023 (DPDP Act); Algorithmic Bias; Financial Cybersecurity.
INTRODUCTION
Artificial Intelligence (AI) is a groundbreaking technology in banking to detect fraud, to control risks, and to become more personalized, but it is also the new means of enhancing advanced criminal activities such as deepfakes and voice cloning to become a representative of an official or a client. Generative AI is used to commit fraudsters who bypass biometric KYC (Know Your Consumer), commit social engineering using LLMs (Large Language Model), and generate synthetic identities, causing massive debt such as the 45 crores in India in 2024 and 2025[1]. Also, new reports found out a significant uprising in AI- enabled financial crime in India. This resulted with a total loss reaching over ₹22,800 crore in 2024 alone[2].Suchattacks destroy customer trust, result in financial distress of identity theft and unauthorized transactions, and flood the detection systems with emerging tactics. Banks experience false negatives which enable fraud slippage, algorithmic biases in identifying legitimate activity, and privacy concerns because of huge sensitive data-sets. Current cyber regulations such as the Information Technology Act 2000 in India are not sufficient to deal with AI-related threats, such as deepfakes, and do not include a basis to hold the accountable in the case of cross-border jurisdiction. The present paper assesses the twofold role of AI in banking practices, regulatory loopholes, ethical issues, and introduces a cohesive system uniting the law, ethics, and technology to reduce risks and promote safe AI implementation.
[1]Volodymyr Zverev et al., Artificial Intelligence and Cybercrime: New Challenges and Prospects for Legal Regulation, 1 Contemp. Issues in Artificial Intelligence 1 (2025).
[2]E. Oztemel& Muhammed Işık, A Systematic Review of Intelligent Systems and Analytic Applications in Credit Card Fraud Detection, 25 Applied Sci. 1 (2025).
Arushi Mehta, Impact of Technological Advancements on Banking Frauds: A Case Study of Indian Banks, 7 Int’l J. Res. Fin. & Mgmt. 261 (2024).