Kumari, D. Jaya and Tejaswi, Gurram and Sri Jahnavi, Nekkanti Durga and Anusha, Korapati and Kathyayani, Kotakonda Naga and Sri, Areti Divya and Sharmila, Medapati (2025) AI-Powered UPI Fraud Detection. International Journal of Innovative Science and Research Technology, 10 (4): 25apr830. pp. 1208-1213. ISSN 2456-2165

[thumbnail of IJISRT25APR830 (1).pdf] Text
IJISRT25APR830 (1).pdf - Published Version

Download (666kB)

Abstract

A significant step in safeguarding India's digital economy against cyber threats is the implementation of AI-driven fraud detection systems in Unified Payments Interface (UPI) transactions. Real-time transaction analysis is made possible by AI technologies, especially machine learning and deep learning, which aid in identifying anomalous patterns that might point to fraud. In 2023, there will be over 100 billion UPI transactions, increasing the need for sophisticated fraud detection techniques. These systems utilize anomaly detection, behavioral biometrics, and network analysis to monitor user interactions and transaction patterns. AI analyzes keystroke dynamics, mouse movements, and transaction history to differentiate legitimate users from fraudsters. Research shows that Generative AI (GenAI) enhances fraud detection accuracy by continuously assessing behavioral patterns, enabling swift identification of suspicious activities. Additionally, combining AI models like Random Forest, Naïve Bayes, and Support Vector Machines (SVMs) improves detection efficiency while reducing false positives. The flexibility of these AI models is crucial for combating new fraud methods, including deepfake scams and synthetic identity fraud. Additionally, initiatives like the Reserve Bank of India's MuleHunter.ai are instrumental in identifying mule accounts involved in illegal transactions and facilitating real-time fraud monitoring among financial institutions. This joint effort strengthens the security infrastructure while ensuring adherence to regulatory requirements for anti-money laundering and counter-terrorism financing. The growing use of AI-driven solutions to identify UPI fraud signifies a notable change in how financial institutions address security issues in an increasingly digital economy. With 72% of financial institutions in India currently employing or considering Generative AI (GenAI)-based technology for fraud prevention, the sector is experiencing a significant transformation that emphasizes the importance of security alongside user experience. As these technologies evolve, they will be vital in fostering consumer confidence and preserving the integrity of India's digital payment landscape in the face of changing cyber threats.

Item Type: Article
Subjects: T Technology > T Technology (General)
Divisions: Faculty of Engineering, Science and Mathematics > School of Engineering Sciences
Depositing User: Editor IJISRT Publication
Date Deposited: 28 Apr 2025 09:51
Last Modified: 28 Apr 2025 09:51
URI: https://eprint.ijisrt.org/id/eprint/587

Actions (login required)

View Item
View Item