Published Paper Details:

The rapid spread of synthetic identities and AI-generated deepfake threats poses significant challenges to traditional enterprise security systems. Modern attack methods use fixed authentication system vulnerabilities to allow criminals to realistically mimic authorized users. The security framework Zero Trust Architecture (ZTA) has emerged as a powerful defensive model because it relies on "never trust, always verify" principles. This study evaluates the potential benefits of AI-based detection systems for improving Zero Trust Architecture's ability to tackle modern threats through adaptive and context-aware protection mechanisms. Centering on continuous authentication, behavioral biometrics, anomaly detection, and advanced identity verification, better identity assurance and fraud prevention are supported. This paper surveys the state-of-the-art methods while examining their existing challenges including model bias and adversarial evasion and scalability issues together with privacy risks and assesses their practical deployment potential. Future research directions presented in this paper focus on uniting AI-based defenses with Zero Trust frameworks to protect identities and sustain digital trust while preventing sophisticated fraud.