The rapid growth of cyber threats has created significant challenges for organisations in protecting digital infrastructures and sensitive information. Traditional cybersecurity systems often fail to provide transparent and understandable decision-making processes, making it difficult for security analysts to interpret automated threat responses. This research proposes an Explainable Artificial Intelligence (XAI)-based intelligent security recommendation system designed for real-time cyber threat mitigation. The framework integrates machine learning, deep learning, and explainable AI techniques to detect, analyse, and recommend appropriate security actions against evolving cyberattacks. Unlike conventional black-box AI models, the proposed system enhances interpretability by providing clear explanations for threat predictions and mitigation strategies. The system collects real-time network traffic and behavioural data, processes them through intelligent threat analysis modules, and generates adaptive security recommendations for intrusion prevention. The proposed model aims to improve cybersecurity efficiency, trustworthiness, transparency, and decision-making accuracy in enterprise environments. Experimental analysis demonstrates that the integration of explainable AI significantly increases user confidence and reduces false-positive security alerts while maintaining high detection accuracy. This study contributes to modern cybersecurity research by presenting a scalable and intelligent framework for proactive cyber defence and secure digital ecosystems.