Beyond Situational Crime Prevention: How Artificial Intelligence Transforms Crime Prevention Theory into Online Retail

Abstract The exponential growth of online retail has been accompanied by a parallel rise in sophisticated cybercrime, which threatens consumer trust and economic stability. Unlike existing research that focuses solely on AI security tools, the current study systematically utilises artificial intelligence (AI) in conjunction with situational crime prevention, identifying important insights from crime prevention studies. While AI has emerged as a powerful technical countermeasure, its application is often disconnected from established criminological theory. This study presents a systematic literature review of AI-based online retail security through the lens of criminological theory, revealing that current technology-based measures neglect theoretical considerations. Following the PRISMA protocol, we searched four major academic databases, namely Web of Science, ACM, IEEE Xplore, and Scopus, for peer-reviewed articles published in the past 15 years. Our analysis of the final corpus of selected studies reveals that, while AI techniques such as anomaly detection, natural language processing, and biometric analysis are increasingly effective against threats such as fraud and account takeovers, their deployment aligns with SCP strategies largely implicitly rather than by design. The findings indicate a strong focus on increasing the risks and efforts for offenders, but significantly less attention is paid to reducing rewards, reducing provocations, or removing excuses. This review highlights critical gaps in the literature: the absence of an integrated framework that integrates AI’s technical capabilities with SCP’s theoretical robustness. By mapping current AI applications to the five pillars of SCP, this paper offers a novel synthesis that bridges the divide between computer science and criminology, proposing a more holistic, theoretically grounded approach to securing the future of online retail.