Blockchain-Driven Federated Learning for Real-Time Threat Defense and Forensic Insight

Real-time, privacy-conserving forensic analysis is needed as cyber threats become more sophisticated. Centralized machine learning models face data privacy, security, and scalability issues, while traditional federated learning does not have integrity verification or adaptive threat indication. This paper presents a Blockchain-Integrated Federated Learning (BFL) framework for supporting digital forensics and real-time cyber threat defense. BFL facilitates forensic organizations to jointly train models without sharing raw data, ensuring data privacy. Blockchain guarantees tamper-evident forensic logs, safe model updates, and automated verification via smart contracts. Experimental outcomes verify that BFL greatly surpasses centralized and traditional federated learning methods, with a detection accuracy of 98.5% for cyber threats. Incorporating blockchain enhances model security, verifies latency, and increases adversarial robustness. By integrating distributed learning with immutable forensic logging, the framework proposed provides an scalable, reliable solution to real-time cyber protection. The research sets an encouraging basis for next-generation AI-based digital forensic systems in secure and smart cybersecurity investigations.