DEVELOPMENT OF A LINUX-BASED UTM SYSTEM USING GENERATIVE ARTIFICIAL INTELLIGENCE AND ENSEMBLE LEARNING TO DETECT URL-BASED PHISHING ATTACKS

The rapid growth of web-based services and digital communication has contributed to the increasing prevalence and sophistication of URL-based phishing attacks. Traditional security mechanisms often struggle to detect newly generated and obfuscated phishing URLs, creating a need for intelligent and adaptive cybersecurity solutions. This study proposes the development of a Linux-based Unified Threat Management (UTM) system that integrates Generative Artificial Intelligence (Generative AI) and ensemble learning techniques for automated detection and classification of phishing URLs. The proposed system combines URL feature extraction, machine-learning-based classification, ensemble decision-making, and AI-assisted analysis within a Linux-based security architecture.