Conceptual Framework for ML-Based Fraud Detection and AML Compliance

Data Quality

Technological Innovation

Regulatory Alignment

Collaboration

Organizational Readiness

Real-Time Monitoring

Model Explainability

Continuous Learning

Secure Data Sharing

Successful Implementation of ML-based System

Improved Detection Accuracy

Reduced False Positives

Enhanced Compliance Efficiency

Increased Customer Trust

Framework derived from six analyzed cases and grounded in Routine Activity Theory (RAT), Risk-Based Approach (RBA), and Technology Acceptance Model (TAM), which collectively explain the behavioral, regulatory, and technological dimensions of ML-based fraud detection and AML systems.