Hybrid Explainable AI, Federated Learning, and Blockchain Framework for Trustworthy Cybersecurity
TL;DR
The proposed hybrid framework offers an effective solution for secure and trustworthy next-generation cybersecurity in IoT, cloud, and enterprise environments and demonstrates improved threat detection performance, enhanced privacy preservation, greater transparency, and stronger resilience against adversarial attacks.
Abstract
The increasing adoption of cloud computing, the Internet of Things (IoT), and distributed networks has intensified cybersecurity challenges, requiring intelligent, secure, and privacy-preserving threat detection mechanisms. This paper proposes a hybrid framework that integrates Explainable Artificial Intelligence (XAI), Federated Learning (FL), and Blockchain to develop a trustworthy cybersecurity system. Federated Learning enables collaborative model training without sharing sensitive data, blockchain ensures secure and tamper-resistant verification of model updates, and XAI techniques such as SHAP and LIME provide transparent explanations for cyber threat predictions. The proposed framework is evaluated using benchmark intrusion detection datasets based on metrics including accuracy, precision, recall, F1-score, blockchain latency, and communication overhead. The results demonstrate improved threat detection performance, enhanced privacy preservation, greater transparency, and stronger resilience against adversarial attacks. The proposed hybrid framework offers an effective solution for secure and trustworthy next-generation cybersecurity in IoT, cloud, and enterprise environments.