AI-POWERED SECURE AUTHENTICATION (BEHAVIOURAL BIOMETRICS)

Abstract

Artificial intelligence (AI) has significantly advanced
biometric authentication by enhancing identification accuracy
and improving system efficiency. Modern biometric systems
increasingly rely on machine learning and deep learning
techniques to strengthen traditional methods such as fingerprint,
facial, iris, and voice recognition. These technologies enable
faster processing and more reliable authentication, making them
widely applicable across various domains.However, the growing
complexity of AI-based systems also introduces important
security challenges. Vulnerabilities may arise during different
stages, including data collection, model training, and system
deployment. These weaknesses can be exploited by attackers
through techniques such as adversarial attacks, data poisoning,
and unauthorized access to sensitive information, leading to
potential privacy violations.This paper analyzes these issues by
reviewing existing research and identifying key gaps in current
approaches. It also suggests the need for developing more
secure, resilient, and privacy-preserving solutions. The primary
objective is to ensure that AI-powered biometric systems
maintain high performance levels while effectively addressing
emerging security threats and safeguarding user data.