AI and IoT driven intelligent risk control and verification system for international trade settlement

Abstract

Cross-border trading and digital financial platforms are growing rapidly, making the settlement of international trade more complex and complex, which requires reliable verification of transactions and effective financial risk control. Traditional rule-based techniques are not flexible in the dynamic trading context. To overcome these limitations, an IoT-powered intelligent verification and AI-driven risk control model is introduced to improve the authenticity of transactions, provide real-time monitoring and improve proactive financial risk control based on distributed sensing technologies and adaptive AI algorithms. The IoT Trade Settlement Risk dataset, comprising 9990 samples, is used to collect transactional records, trade documentation metadata and payment behaviour data. Data preparation involves applying Min–Max normalization and Savitzky–Golay filtering to make data consistent from different sources. The discriminative information is represented by Principal Component Analysis (PCA) of critical trade indicators. The core analytical engine employs a Termite Alate Optimization-driven Adaptive Deep Convolutional Neural Network (TAO-ADCNN) model perform multi-class transaction risk classification by categorizing each trade transaction into low-, medium-, or high-risk classes. The predicted risk category is subsequently utilized to support intelligent transaction verification and anomaly monitoring. Convolutional parameters and network weights to improve convergence efficiency and global search capability, while the ADCNN analyzes hierarchical transaction patterns for accurate risk categorization. Experiments using Python have shown that the suggested TAO-ADCNN model outperforms the other models and datasets, attaining higher accuracy results, with a range of 96% to 98.9%. The overall results obtained the identification of risks, verification, and resilience are improved in international trade settlement management systems. TAO-ADCNN enables proactive monitoring of financial and operational risks. System improves classification consistency across varied trading scenarios. Efficient risk detection allows early action for high-risk transactions. System enhances transparency and accountability in cross-border trade systems.