This research introduces a hybrid framework combining Autoencoders and Large Language Models to enhance IoT security through human-comprehensible insights.

• Integrates numerical anomaly detection with language models to interpret complex IoT data
• Improves preprocessing of heterogeneous IoT data streams
• Enables security analysts to understand threats through natural language explanations
• Bridges the gap between machine detection and human decision-making

This innovation significantly improves security posture for critical infrastructure by transforming cryptic security alerts into actionable intelligence, allowing faster response to potential threats and reducing the expertise barrier for IoT security.

Enhancing Cybersecurity in Critical Infrastructure with LLM-Assisted Explainable IoT Systems