AI-BASED GST FRAUD DETECTION SYSTEM FOR LARGE-SCALE E-COMMERCE PLATFORMS

The rapid growth of e-commerce platforms has multiplied shall we say the fraudulent activities related to Goods and Service Tax (GST). Due to large scale as well as evolving fraud patterns, traditional fraud detection mechanisms are unable to scale and effectively detect fraud. To identify such hidden relationships between entities (sellers, transactions and their tax records), this research presents an AI based GST fraud detection system employing Graph Neural Networks (GNNs). The system uses Google Cloud AI to deploy the system on the scale to handle processing of huge amounts of transactional data, spotting anomalies, and alerting on potential fraud real time. The proposed solution increases fraud detection accuracy while decreasing false positives using different patterns of tax evasion. We conduct experiments for large scale e-commerce ecosystem and our approach is found to be effective in identifying fraudulent activities in such scenarios. Keywords: GST Fraud Detection, Graph Neural Networks (GNNs), E-Commerce Security, Anomaly Detection, Google Cloud AI, Large-Scale Tax Compliance