Explainable AI (XAI) for Cybersecurity Decision-Making in Industrial Automation
Pawan Kumar Goel, Chin‐Shiuh Shieh, Mong‐Fong Horng · Advances in computational intelligence and robotics book series · 2025
The integration of Artificial Intelligence (AI) in cybersecurity for industrial automation is crucial due to the increasing reliance on smart technologies and the sophistication of cyber threats. This chapter focuses on the application of Explainable AI (XAI) in cybersecurity decision-making, addressing the challenges of interpreting and trusting AI-driven security solutions in complex industrial environments. Existing literature highlights issues like lack of transparency in AI models, difficulty in real-time threat interpretation, and the need for human-understandable explanations. To address these, a novel XAI framework is proposed, combining interpretable machine learning models with domain-specific knowledge to provide actionable and transparent cybersecurity insights. Experimental results show significant improvements in threat detection accuracy, interpretability, and response times, achieving a 20% increase in detection rates and a 30% reduction in false positives.