Generative AI–Guided Sentinel for Self-Optimizing Federated Cybersecurity and Intelligent Threat Detection
Akter Rokaya, Md Al Samiul Amin Rishat, Singh Sudhanshu, Abhishank Singh, Bian Naizheng, Akter Rokaya, Md Al Samiul Amin Rishat, Singh Sudhanshu, Abhishank Singh, Bian Naizheng · Journal of Hunan University Natural Sciences · 2025
As cyber threats become increasingly sophisticated and pervasive, adaptive, intelligent, and privacy-preserving intrusion detection systems (IDSs) are more critical than ever, particularly in ecosystem-based networks. However, existing federated cybersecurity systems continue to face several challenges, including reduced data processing efficiency caused by data heterogeneity, difficulties in real-time threat detection, and complex configuration management. To address these challenges, we propose a novel Self-Improving Federated Cybersecurity Sentinel framework that integrates Federated Learning (FL) with Generative Artificial Intelligence (GAI) to enable dynamic and context-aware optimization. Large Language Models (LLMs) play a central role in the proposed framework by enabling prompt-driven analytics for real-time intrusion detection, feature relevance assessment, automated anomaly investigation, and adaptive optimization of FL parameters. Experimental evaluations conducted on the UNSW-NB15 dataset demonstrate that the proposed framework achieves a precision of 0.96 and an F1-score of 0.98, while simultaneously reducing configuration adjustment time by approximately 68%. These results indicate that the framework significantly simplifies the tuning process and enhances detection performance. Overall, this study represents a substantial advancement toward fully autonomous and robust cybersecurity systems. Future work will focus on improving the generalizability of the framework across diverse threat scenarios and incorporating explainable AI components to enhance transparency and interpretability. Keywords: Federated Learning, Large Language Model, Cybersecurity, Intrusion Detection, Generative AI.