AI in Digital Payments: Security and Fraud Detection

AI and Machine Learning have now become the key technologies to enhance the security of digital payment systems and detect any instances of fraud. The literature review reveals that both supervised learning, unsupervised anomaly detection, deep learning, and hybrid AI systems can be used to detect suspicious behaviour and enable quicker fraud detection. In comparison to the traditionally used rule-based systems, the use of AI is more flexible when it comes to the analysis of large amounts of transaction data. Various types of AI bring varying strengths when it comes to the detection of frauds. Examples of supervised models include Logistic Regression, Random Forest, Decision Tree, and XGBoost which are helpful when it comes to detecting fraud cases that are already known. Unsupervised learning techniques like Isolation Forest and Autoencoders help in detecting abnormal behavior of transactions. The hybrid approach combines all these techniques to enhance the detection of known and emerging fraud cases. On the other hand, some limitations are presented by the selected studies when it comes to the application of AI to enhance security in digital payments. This is because the reviewed studies have indicated the limitations of unbalanced datasets, the absence of labelled fraud data, high false-positive rates, privacy problems, nonexplainable nature, computational demands, regulation, and constant changes in fraud schemes. Future research could consider working on creating adaptive hybrid systems, increasing explainability of AI, improving methods that protect privacy, and developing real-time fraud detection tools. Methods such as federated learning and novel AI systems may help to enhance the security of payments without compromising data privacy Apart from developing technologies, proper security procedures, regulatory compliance, and user knowledge are also vital. When we pull this all together, AI holds real promise for making digital payments more secure, more reliable, and more efficient. Intelligent fraud detection, explainable decisions, privacy protection, monitoring, and responsible deployment together will make it far easier for financial institutions to stay ahead of fraud and keep digital payments secure enough that users can genuinely rely on them.

[1]  J. Bari, R. Pawar, I. Holkar, and R. Dekate, “Strengthening Digital Payment and Security using AI, Big Data, and Cybersecurity in India,” Advanced International Journal for Research, vol. 7, no. 2, 2026. https://doi.org/10.63363/aijfr.2026.v07i02.4814 [2]   D. Dahiphale et al., “Enhancing Trust and Safety in Digital Payments: An LLM-Powered Approach,” 2024. https://doi.org/10.48550/arXiv.2410.19845  [3]  A. B. Lahari, K. K. Kumar, B. Neelima, K. Raghu, and K. P. Varsh, “Digital Payment Fraud Detection Using Machine Learning,” International Journal for Multidisciplinary Research, vol. 8, no. 2, 2026. https://www.ijfmr.com/researchpaper.php?id=73671 [4]   V. Chang et al., “Digital payment fraud detection methods in digital ages and Industry 4.0,” Computers and Electrical Engineering, vol. 100, Art. No. 107734, 2022. https://doi.org/10.1016/j.compeleceng.2022.107734  [5]   M. Sorinola and K. Hamed, “Machine Learning for Digital  Payment  Fraud  Detection  in Financial Institutions: A Review of Methods, Explainability, and Performance,” Cureus Journal of Computer         Science,  2026. https://doi.org/10.7759/s44389-026-00149-0  [6]   L. Koduru, “Harnessing AI to Enhance Security in Digital Payments: Detection of High-Risk Users and Mitigation of Financial Crime,” International Journal of Intelligent Systems and Applications in Engineering, vol. 12,   no.          17s,        pp.          918–927,       2024. https://www.ijisae.org/index.php/IJISAE/article/view/76 26 [7]  A. Shivarudraiah, “AI-Powered Threat Detection in Digital Payments: Addressing Cyber Fraud,”International Journal of Artificial Intelligence, Data Science, and Machine Learning, vol. 3, no. 4, pp. 19–26, 2022. https://doi.org/10.63282/3050-9262.IJAIDSML- V3I4P103[8]  S. M. S. Shaikh and G. N. A, “AI Based Transaction Anomaly and Fraud Detection System,” International Research Journal on Advanced Engineering Hub, vol. 4, no.  3,      pp.        1160–1169,               2026.https://doi.org/10.47392/IRJAEH.2026.0164 [9]  N. B. Chakka and S. S. Saheb, “A Hybrid Autoencoder-XGBoost Framework for High-Performance UPI Fraud Detection,” Proceedings of Engineering and Technology Innovation, vol. 33, pp. 93–105,  2026.https://doi.org/10.46604/peti.2026.16002  [10]A. Davitaia, “Artificial Intelligence and machine learning in fraud detection for digital payments,” International Journal of Science and Research Archive, vol.   15,   no.   3,   pp.   714–719,   2025. https://doi.org/10.30574/ijsra.2025.15.3.1784