With the increasing prevalence of sophisticated malware, ensuring robust cybersecurity has become more critical than ever. Traditional detection methods—such as static and dynamic analysis—face limitations in identifying complex and evolving threats. This research presents a hybrid malware detection approach that integrates static and dynamic analysis with AI-based classification to improve detection accuracy and system resilience. Key features are extracted from APK and DLL files to train machine learning models capable of distinguishing malicious files with greater precision. The proposed system generates detailed detection reports, quarantines infected files, and maintains logs for iterative improvement. Beyond enhancing detection capabilities, the system reduces redundant scanning operations and computational overhead, contributing to energy-efficient computing. By minimizing resource usage and false positives, this approach supports the creation of sustainable and resilient cybersecurity infrastructures. The study also addresses challenges in AI-driven malware detection and outlines future directions for advancing digital security within the framework of long-term sustainability goals.