PROACTIVE DETECTION OF ZERO-CLICK ATTACKS USING STATIC ANALYSIS AND SANDBOX-BASED BEHAVIORAL MONITORING

Abstract

Zero-click attacks represent a critical threat to modern digital communication systems by exploiting vulnerabilities in message parsing engines without requiring any user interaction. These attacks, exemplified by spyware like Pegasus, operate silently and often evade traditional detection mechanisms that rely on user actions or known malware signatures. This research presents a multi-layered detection framework that proactively mitigates zero-click threats using a combination of static payload analysis and sandbox-based behavioral inspection.The proposed solution employs an OS-level pre-parser to identify anomalous file structures, headers, and metadata in incoming messages, followed by dynamic analysis in a secure sandbox environment. Evaluation through simulated Pegasus-like payloads and benchmarked comparisons with conventional antivirus and intrusion detection systems demonstrated a detection accuracy of 95.1%, with a significant reduction in false positives to 4.8%. Performance remained within acceptable limits for real-time environments, with minimal processing overhead.This approach effectively stops malicious payloads before execution, adheres to Zero Trust principles, and functions independently of user behavior or delayed patching cycles. Future enhancements include the integration of adaptive machine learning models, improved handling of encrypted data streams, and scalable deployment across mobile OS architectures and enterprise gateways. This work offers a proactive, agnostic, and scalable defense mechanism against one of the most sophisticated cyberattack vectors in modern threat landscapes.