Cyber-Physical Systems (CPS) are rapidly evolving in every Critical Infrastructure (CI) domain. While the custodians of such systems continuously endeavour to protect their infrastructure, the threat always looms as the opponents refine their tactics and attack methodologies. This research aims to investigate these issues by developing a technique(s) that would allow CI service providers to identify malware threats facing their network in an automated, real-time manner, ultimately providing actionable intelligence to enable the mitigation of these threats. This investigation will first distinguish between CPS and non-CPS malware by analysing different sample sets through feature selection/extraction. Once classified, the research will employ different nature-inspired algorithms designed for bringing efficiency through fast processing and improved accuracy by reducing false positives. The proposed mechanism is expected to benefit CI sectors such as the telecommunication providers, automated/self-driving cars network, financial systems, electricity/energy, water by helping identify cyber threats through artificial intelligence. Through this, early detection of malware and their appropriate mitigation would become possible, thereby protecting the organisation’s critical network infrastructure and proprietary data. The research has been envisioned as industry-driven that will help solve actual problems for a nation’s community, government, and industry.