AI Driven Fraud Detection in FinTech Using Advanced Machine Learning and Encrypted Data Pipelines

Abstract

Financial technology (FinTech) systems are increasingly targeted by sophisticated fraud schemes that exploit digital transaction pathways. Traditional rule-based fraud detection systems lack adaptability and often produce high false positive rates. This paper presents an integrated framework combining advanced machine learning (ML) models with encrypted data pipelines to improve fraud detection accuracy while preserving data security and privacy. The research evaluates various data encryption mechanisms, ML algorithms, and their performance within real-world FinTech datasets. Results demonstrate that encrypted pipeline architectures combined with deep learning and anomaly detection techniques significantly enhance fraud detection effectiveness with minimal computational overhead.