Year: 2025
Type: article
Abstract: The rapid advancement of digital payment systems has transformed the way financial transactions are conducted, providing users with unmatched speed, convenience, and accessibility on a global scale.However, this progression has also made digital platforms increasingly vulnerable to sophisticated fraudulent activities.This study investigates how the integration of Artificial Intelligence (AI) and Machine Learning (ML) can enhance the mechanisms for detecting fraud within digital payment ecosystems.The research outlines a comprehensive fraud detection workflow that includes key stages such as data collection, preprocessing, feature engineering, model training, real-time inference, and adaptive learning.It includes case studies from various sectors, including financial institutions and telecommunications companies like 3UK, showcasing practical applications and effectiveness in mitigating fraud.Additionally, the study assesses crucial ML techniques such as supervised learning, anomaly detection, and reinforcement learning, while addressing challenges like data imbalance, the evolution of fraud patterns, and explainability of the models used.The findings indicate that AI significantly improves both the accuracy and speed of fraud detection efforts, leading to reduced financial losses and enhanced consumer trust.The paper concludes by highlighting future prospects such as the implementation of federated learning, behavioral biometrics, and blockchain-based AI models, all aimed at fostering more secure, scalable, and privacy-preserving fraud prevention strategies in digital payments.
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4409644541