Abstract

This chapter analyzes the transformation of social engineering through artificial intelligence, tracing its evolution from historically manual deception practices to algorithmically optimized, multimodal manipulation systems. It first situates contemporary AI-enabled attacks within the historical lineage of psychological exploitation, emphasizing continuity in cognitive bias targeting despite technological escalation. The discussion then examines how machine learning enhances phishing, vishing, and smishing campaigns through behavioral profiling, adaptive messaging, timing optimization, and cross-channel orchestration. Focus is given to deepfake technologies and voice cloning systems, outlining their technical foundations, including GAN-based visual synthesis, neural speech replication, and multimodal synchronization frameworks. The chapter further explores synthetic identity fraud mechanisms, where fabricated personas integrate cloned voices, generated imagery, and automated conversational agents to bypass verification infrastructures.