Abstract
Email is a ubiquitous communication tool. This ubiquity has made emails a frequent target of phishing attacks. Organizations often use digital nudges (interventions designed to guide user behavior) to warn users of a phishing email. However, these warnings are often uninterpretable and thus often ignored. Therefore, this study investigates how explainable artificial intelligence (XAI) can be implemented in a digital nudge to enhance user phishing detection and prevention. In this research, we designed an XAI-enabled nudge and conducted a between subjects’ experiment comparing it against a popular extant nudge and a control group of no nudge. The results suggested that XAI-enabled nudges significantly improve user phishing detection accuracy by 10.1%. Additionally, XAI-enabled nudges significantly decreased intrinsic and germane cognitive load. These findings contribute to XAI research by demonstrating the value of AI explanations in security contexts and its impact on cognitive processing. These results also have practical implications for email security system integration design.
Recommended Citation
Wagner, Mason; Ampel, Benjamin M.; Hashim, Matthew; and Chen, Hsinchun, "Email Phishing Prevention: An Explainable Nudging Approach" (2025). WISP 2025 Proceedings. 4.
https://aisel.aisnet.org/wisp2025/4