Integrating Blockchain and Machine Learning for Predictive Cyber Defense Systems

Kohistani, Ahmad Jamy
Azimi, Irfanullah
Fazil, Abdul Wajid

ARMADA : Jurnal Penelitian Multidisiplin
Vol. 3 No. 12 (2025): ARMADA : Jurnal Penelitian Multidisplin, December 2025

The rapid expansion of cyber threats targeting critical infrastructures highlights the limitations of traditional centralized security systems, which suffer from latency, scalability constraints, and single points of failure. This study addresses this problem by examining how the integration of Blockchain and Machine Learning (ML) can strengthen predictive cyber defense and enhance real-time anomaly detection. The purpose of the research is to synthesize current evidence on the security, efficiency, and operational benefits of Blockchain–ML frameworks through a Systematic Literature Review (SLR). Following PRISMA guidelines, a structured search was conducted across four major databases IEEE Xplore, ScienceDirect, Scopus, and Web of Science covering peer-reviewed literature published between 2020 and 2025. Using a four-category keyword strategy, the review initially identified 1100 records, ultimately narrowing the final dataset to 25 studies that met all inclusion criteria. The results indicate that Blockchain significantly enhances data integrity, auditability, and threat-intelligence reliability, while ML improves predictive accuracy and supports real-time detection. Together, these technologies outperform conventional centralized systems in terms of transparency, resilience, and operational efficiency. The study concludes that Blockchain–ML integration provides a robust foundation for next-generation, decentralized cybersecurity architectures, offering measurable improvements in security and system performance.