Identity-Backed Biometric Authentication: a Zero-Trust Framework for Defeating Synthetic Identity Attacks in Regulated Financial Institutions

Abstract

Synthetic identity fraud and deepfake-based impersonation attacks have become major cybersecurity threats in regulated financial institutions due to the rapid growth of digital banking and AI-generated identity manipulation techniques. Many traditional authentication methods like password authentication, token authentication, and standalone biometric methods are ineffective against advanced spoofing attacks, namely replay attacks and advanced synthetic identities. To address these issues, this paper introduces a novel framework for secure and continuous identity verification called Identity-Backed Biometric Authentication (IBBA) that adheres to the Zero-Trust security principles for a financial environment. The proposed framework combines facial biometric authentication, liveness detection, identity consistency validation, behavioral analytics, adaptive trust evaluation, and anomaly detection using machine learning to constantly check the legitimacy of users during the transaction process. This framework actively monitors biometric confidence, behavioral consistency, device trust, and transaction risk to identify suspicious activity and combat unauthorized access. The experimental results demonstrate that the proposed authentication framework, IBBA, achieves 97.8% authentication accuracy, outperforms traditional authentication methods in resisting deepfake-based impersonation attacks, replay attacks, and synthetic identity fraud, and boosts the authentication system's fraud detection capability. The findings from this study also support the benefits of the proposed zero-trust biometric approach, such as bolstering transaction security, identity verification, and regulatory compliance, and bolstering cybersecurity resilience in the contemporary digital financial landscape.