Abstract
Social engineering remains a leading pathway to cyber harm, not primarily because users “lack awareness,” but because attackers exploit common cognitive shortcuts, social norms, and organisational routines – now intensified by rapid, platform-mediated communication. This study aims to clarify why social engineering persists as a prevention challenge and to assess the limitations of current prevention strategies. The paper employs a narrative (critical) literature review. Interdisciplinary research on social engineering and cybercrime prevention is synthesised using a thematic-analytic structure across three strands: (1) social vulnerability and victimisation, (2) educational and organisational prevention (training, awareness, culture, routines, and usable warning/interface design), and (3) technological asymmetry and escalation, including AI-enabled deception and deepfake-enabled impersonation. The reviewed literature converges on four findings. First, susceptibility is contextual and patterned, shaped by roles, routines, socio-emotional dynamics, and platform cues rather than stable individual deficits. Second, training and awareness can help but often show limited transfer and sustainability when gener