Cloud-Native Multi-Factor Authentication Framework for Digital Banking Systems: An AI-Driven Adaptive Security Architecture

Abstract

The digital banking ecosystem faces unprecedented security challenges with the imminent advent of quantum computing, which threatens to compromise existing cryptographic infrastructures. Traditional multi-factor authentication (MFA) systems in banking rely on classical cryptographic primitives vulnerable to quantum attacks, creating critical security gaps in financial transactions and customer data protection. This paper introduces a novel Quantum-Enhanced Multi-Factor Authentication Framework (QE-MFAF) specifically designed for digital banking systems, integrating lattice-based post-quantum cryptography with biometric fuzzy commitment schemes and behavioral analytics. The proposed framework incorporates a hybrid authentication mechanism combining quantum-resistant cryptographic protocols, continuous behavioral pattern recognition using deep learning models, and secure multi-party computation for transaction verification. Experimental validation on a synthetic banking dataset demonstrates superior performance with 99.7% authentication accuracy, 0.02% false positive rate, and 15ms average authentication latency while maintaining quantum resistance. The framework successfully mitigates advanced persistent threats, insider attacks, and quantum-based cryptographic vulnerabilities while ensuring seamless user experience in high-transaction banking environments. Performance comparisons with existing banking authentication systems show 34% improvement in security metrics and 28% reduction in computational overhead.