Effectiveness of Artificial Intelligence-Based Adaptive Honeypots in Cyber Threat Detection: A Systematic Literature Review and Meta-Analysis

This study conducted a systematic literature review and meta-analysis on using honeypot systems enhanced by artificial intelligence to improve the effectiveness of cyber threat detection in organizational environments. The review followed the PRISMA protocol and assessed 454 articles from databases, including IEEE, ACM, Emerald, and Web of Science. After a multi-stage screening process, 62 articles met the inclusion criteria and were further analyzed. The synthesis indicated that integrating artificial intelligence into honeypot systems improved detection accuracy, expanded the system’s ability to recognize varied attack patterns, and optimized resource efficiency. A meta-analysis of 36 studies revealed consistent, significant improvements in detection performance. Quantitatively, the analysis yielded a mean effect size of 0.905, indicating a substantial improvement in detection effectiveness resulting from AI integration. These findings confirm that adopting AI-based honeypot technologies is essential for addressing increasingly complex cyberattacks and for providing a foundation for future research into the development of standardized evaluation frameworks.