Hybrid Explainable AI, Federated Learning, and Blockchain Framework for Trustworthy Cybersecurity

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.

Introduction

The study proposes a hybrid cybersecurity framework that integrates Explainable Artificial Intelligence (XAI), Federated Learning (FL), and Blockchain to provide secure, privacy-preserving, and trustworthy cyber threat detection in distributed environments such as cloud computing, IoT, edge computing, and enterprise networks.

Traditional AI-based cybersecurity systems rely on centralized data collection, which raises privacy and security concerns. To overcome this limitation, the proposed framework uses Federated Learning, allowing multiple organizations to collaboratively train intrusion detection models without sharing raw data. However, since FL is vulnerable to malicious updates and poisoning attacks, Blockchain is incorporated to securely validate and record model updates using smart contracts, ensuring integrity, traceability, and decentralized trust. To improve transparency and user confidence, Explainable AI techniques such as SHAP and LIME are applied to explain why the AI classifies network activities as malicious or benign.

The study aims to develop a secure hybrid framework, implement privacy-preserving federated learning, integrate blockchain for secure model verification, apply explainability techniques, and evaluate the system using benchmark datasets including CICIDS2017, UNSW-NB15, and BoT-IoT. The framework follows a layered architecture consisting of distributed data sources, federated learning, blockchain verification, explainable AI, and a real-time cybersecurity application layer. It protects against threats such as data poisoning, model poisoning, Byzantine attacks, tampering, and unauthorized updates.

Experimental evaluation demonstrates that the proposed hybrid framework outperforms conventional AI and standard federated learning models. It achieved 98.7% accuracy, 98.5% precision, 98.3% recall, and an F1-score of 98.4%, exceeding the performance of centralized CNN and traditional FL models. Blockchain verification achieved a 99.8% model verification success rate with only 1.5 seconds transaction latency and low communication overhead, confirming efficient and secure model aggregation. The hybrid XAI approach provided high-quality explanations with fast interpretation, enabling analysts to better understand AI decisions, reduce false positives, and improve incident response.

Overall, the results indicate that integrating Explainable AI, Federated Learning, and Blockchain creates a robust cybersecurity framework that enhances detection accuracy, preserves data privacy, strengthens resistance against adversarial attacks, increases transparency, and establishes trust among collaborating organizations. The framework is suitable for deployment in next-generation distributed cybersecurity environments and provides a scalable foundation for future secure AI-driven cyber defense systems.

Conclusion

This paper presented a hybrid cybersecurity framework that integrates Explainable Artificial Intelligence (XAI), Federated Learning (FL), and Blockchain to provide a secure, privacy-preserving, and trustworthy cyber threat detection system. The proposed framework combines distributed model training through federated learning, secure model verification using blockchain, and transparent decision-making through XAI techniques such as SHAP and LIME. The experimental results demonstrated that the proposed framework achieved high intrusion detection performance with an accuracy of 98.7%, while preserving data privacy and ensuring the integrity of model updates. The blockchain layer enhanced trust by securely validating federated model updates through smart contracts, whereas the XAI module improved transparency by providing interpretable explanations for cyber threat predictions. These features enabled more reliable and informed decision-making for cybersecurity analysts. Overall, the proposed hybrid framework successfully addresses the limitations of conventional AI-based cybersecurity systems by combining privacy preservation, decentralized trust, and explainable intelligence within a unified architecture. The framework is suitable for deployment in cloud computing, Internet of Things (IoT), edge computing, and enterprise network environments. Future work will focus on optimizing communication efficiency, improving scalability, integrating advanced federated learning algorithms, and extending the framework to support real-time detection of emerging cyber threats in large-scale distributed systems.

References

[1] Tjoa, E., & Guan, C. (2021). A survey on explainable artificial intelligence (XAI): Toward medical XAI. IEEE Transactions on Neural Networks and Learning Systems, 32(11), 4793–4813. [2] Zhang, Y., Chen, X., Jin, L., Wang, X., & Guo, D. (2021). Explainable artificial intelligence applications in cybersecurity: State-of-the-art in research. IEEE Access, 9, 102466–102492. [3] Reynaud, Stéphane & Roxin, Ana. (2025). Review of eXplainable artificial intelligence for cybersecurity systems. Discover Artificial Intelligence. 5. 10.1007/s44163-025-00318-5. [4] Muia, Charles & Kamiri, Jackson. (2025). Explainable Artificial Intelligence: A Comprehensive Review of Techniques, Applications, and Emerging Trends. International Journal of Scientific Research in Computer Science and Engineering. 13. 57-68. 10.26438/ijsrcse.v13i4.740. [5] M. Alazab, S. P. RM, P. M, P. K. R. Maddikunta, T. R. Gadekallu and Q. -V. Pham, \"Federated Learning for Cybersecurity: Concepts, Challenges, and Future Directions,\" in IEEE Transactions on Industrial Informatics, vol. 18, no. 5, pp. 3501-3509, May 2022, doi: 10.1109/TII.2021.3119038. [6] Timofte, E. M., Dimian, M., Graur, A., Potorac, A. D., Balan, D., Croitoru, I., Hri?can, D.-F., & Pu?ca?u, M. (2025). Federated Learning for Cybersecurity: A Privacy-Preserving Approach. Applied Sciences, 15(12), 6878. https://doi.org/10.3390/app15126878 [7] Almarri, S., & Aljughaiman, A. (2024). Blockchain Technology for IoT Security and Trust: A Comprehensive SLR. Sustainability, 16(23), 10177. https://doi.org/10.3390/su162310177 [8] Ahmed M. Shamsan Saleh, Blockchain for secure and decentralized artificial intelligence in cybersecurity: A comprehensive review, Blockchain: Research and Applications,Volume 5, Issue 3, 2024,100193,ISSN 2096-7209, https://doi.org/10.1016/j.bcra.2024.100193.

Copyright

Copyright © 2026 Dr. Ravi Chinkera , Dr. Syed Gilani Pasha . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.