Citation:

Deepesh Vinodkumar Semlani, Sudha Rani Pujari , 2025. "AI-Based Fraud Detection in Accounts Payable and Payment Automation Systems", ESP Journal of Engineering & Technology Advancements  5(3): 25-31.

Abstract:

With growing sophistication and numbers of digital finance systems, the risk of fraud in accounts payable (AP) and automated payment systems has deepened. Legacy rule-based controls cannot cope with detecting sophisticated threats like invoice forgery, vendor impersonation, and duplicate disbursements. This review examines the combination of machine learning (ML) and AI with AP processes, showing how smart systems enhance anomaly detection, reduce false positives, and enable scalable fraud protection. It gives best AI techniques, including deep learning, unsupervised outlier detection, explainable AI (XAI), and federated learning. With system designs, theoretical models, and experimental standards, the review here shows the growing maturity and operational value of AI in modern finance. The paper concludes by proposing next-level research, governance models, and technology integration to create trustworthy and preventive fraud detection systems.

References:

[1] PwC. (2022). Global Economic Crime and Fraud Survey: Navigating the rising tide of fraud. PwC Insights. Retrieved from https://www.pwc.com

[2] Association of Certified Fraud Examiners (ACFE). (2022). Report to the Nations: Global Study on Occupational Fraud and Abuse. ACFE. Retrieved from https://www.acfe.com

[3] Gartner. (2023). Market Guide for Accounts Payable Invoice Automation Solutions. Gartner Reports. Retrieved from https://www.gartner.com

[4] Lee, J., & Park, H. (2021). Deep learning-based fraud detection in accounts payable processes. Journal of Financial Crime, 28(4), 1002–1018.

[5] Oracle Corporation. (2023). AI in Oracle Fusion Cloud Financials: Automated Controls and Anomaly Detection. Oracle White Paper. Retrieved from https://www.oracle.com

[6] Lundberg, S. M., & Lee, S. I. (2017). A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 30, 4765–4774.

[7] IBM Research. (2022). Cognitive analytics for payment fraud: Designing robust and scalable systems. IBM Technical Brief. Retrieved from https://research.ibm.com

[8] Chen, L., & Zhou, J. (2018). Detecting fraudulent invoices using machine learning techniques. Journal of Financial Crime, 25(3), 835–846.

[9] Banerjee, A., & Gupta, S. (2019). DeepPay: Deep learning framework for fraud detection in payment automation. Expert Systems with Applications, 127, 248–259.

[10] Li, H., & Yang, Q. (2020). NLP-driven entity resolution in vendor fraud detection. IEEE Transactions on Knowledge and Data Engineering, 32(9), 1827–1839.

[11] Wu, Y., & Zhang, Y. (2020). Graph-based fraud detection in financial transaction systems. Information Sciences, 522, 210–225.

[12] Tang, J., & Lee, D. (2021). A hybrid AI system for detecting duplicate payments in ERP systems. Computers in Industry, 129, 103443.

[13] Sharma, A., & Dey, L. (2021). SHAP-based explanations for financial fraud detection in accounts payable. Computers & Security, 108, 102373.

[14] Patel, V., & Singh, R. (2022). Unsupervised anomaly detection for high-risk AP payment patterns. Pattern Recognition Letters, 155, 30–39.

[15] Ibrahim, R., & Kim, Y. (2022). Cross-ledger analytics for detecting suspicious accounts payable activity. Journal of Accounting and Information Systems, 38(2), 190–205.

[16] Tan, Y., & Zhang, F. (2023). Federated learning for AP fraud detection across enterprise sites. IEEE Transactions on Neural Networks and Learning Systems, 34(4), 1752–1764.

[17] Oracle Corporation. (2023). AI integration for real-time fraud detection in Fusion and SAP ERP. Oracle and SAP Joint White Paper. Retrieved from https://www.oracle.com

[18] IBM Research. (2022). Cognitive analytics for payment fraud: Designing robust and scalable systems. IBM Technical Brief. Retrieved from https://research.ibm.com

[19] Patel, V., & Singh, R. (2022). Unsupervised anomaly detection for high-risk AP payment patterns. Pattern Recognition Letters, 155, 30–39.

[20] Wu, Y., & Zhang, Y. (2020). Graph-based fraud detection in financial transaction systems. Information Sciences, 522, 210–225.

[21] European Commission. (2023). Artificial Intelligence Act: Proposal for a Regulation of the European Parliament. Retrieved from https://digital-strategy.ec.europa.eu

[22] Oracle Corporation. (2023). AI in Oracle Fusion Cloud Financials: Automated Controls and Anomaly Detection. Oracle White Paper. Retrieved from https://www.oracle.com

[23] Gartner. (2023). Market Guide for Accounts Payable Invoice Automation Solutions. Gartner Reports. Retrieved from https://www.gartner.com

[24] Sharma, A., & Dey, L. (2021). SHAP-based explanations for financial fraud detection in accounts payable. Computers & Security, 108, 102373.

[25] Banerjee, A., & Gupta, S. (2019). DeepPay: Deep learning framework for fraud detection in payment automation. Expert Systems with Applications, 127, 248–259.

[26] Tang, J., & Lee, D. (2021). A hybrid AI system for detecting duplicate payments in ERP systems. Computers in Industry, 129, 103443.

[27] Li, H., & Yang, Q. (2020). NLP-driven entity resolution in vendor fraud detection. IEEE Transactions on Knowledge and Data Engineering, 32(9), 1827–1839.

[28] Oracle Corporation. (2023). AI in Oracle Fusion Cloud Financials: Automated Controls and Anomaly Detection. Oracle White Paper. Retrieved from https://www.oracle.com

[29] Sharma, A., & Dey, L. (2021). SHAP-based explanations for financial fraud detection in accounts payable. Computers & Security, 108, 102373.

[30] Banerjee, A., & Gupta, S. (2019). DeepPay: Deep learning framework for fraud detection in payment automation. Expert Systems with Applications, 127, 248–259.

[31] Sharma, A., & Dey, L. (2021). SHAP-based explanations for financial fraud detection in accounts payable. Computers & Security, 108, 102373.

[32] Gartner. (2023). Market Guide for Accounts Payable Invoice Automation Solutions. Gartner Reports. Retrieved from https://www.gartner.com

[33] Tan, Y., & Zhang, F. (2023). Federated learning for AP fraud detection across enterprise sites. IEEE Transactions on Neural Networks and Learning Systems, 34(4), 1752–1764.

[34] Wu, Y., & Zhang, Y. (2020). Graph-based fraud detection in financial transaction systems. Information Sciences, 522, 210–225.

[35] IBM Research. (2022). AI-enabled continuous auditing and anomaly detection in finance. IBM Finance Analytics Report. Retrieved from https://research.ibm.com

[36] European Commission. (2023). Artificial Intelligence Act: Proposal for a Regulation of the European Parliament. Retrieved from https://digital-strategy.ec.europa.eu

[37] Li, Q., Zhang, J., & Zhu, H. (2021). Generating synthetic financial fraud data using GANs for robust model training. Journal of Financial Data Science, 3(2), 54–67.

Keywords:

Accounts Payable, Fraud Detection, Machine Learning, Payment Automation, Invoice Fraud, Anomaly Detection, Deep Learning, XAI, Federated Learning, AP Controls, ERP Security.