Automated Malware Analysis Using AI-Driven Behavioral Analysis Techniques

Abstract

Malware is evolving at an unprecedented rate, bypassing traditional signature-based detection
methods and rendering conventional antivirus solutions ineffective against zero-day threats. This
paper proposes an AI-driven behavioral malware detection framework that employs machine
learning (ML) models, deep learning (DL) architectures, sandbox-based execution, heuristic
analysis, and dynamic malware profiling to enhance threat detection. The proposed system
automates feature extraction, anomaly detection, and classification of malicious files and
executables in real time. The results of extensive experiments conducted on real-world malware
datasets indicate that the proposed model significantly improves detection accuracy, reduces false
positives, and efficiently identifies novel malware strains.