Enhancing IoT Security: Continuous Authentication Through Behavioral Biometrics and AI

This review paper examines the role of behavioral biometrics and artificial intelligence in strengthening continuous authentication for Internet of Things (IoT) systems. It discusses how behavioral signals such as gait, voice, keystroke dynamics, and touch or swipe patterns can support passive and repeated user verification throughout a session. The paper reviews current AI-driven biometric approaches, compares their suitability for IoT security, and highlights key challenges including spoofing attacks, resource constraints, privacy risks, and biometric template protection. It also identifies future research directions such as federated learning, lightweight on-device AI, and stronger anti-spoofing methods.