AN AUTOMATED SCENARIO GENERATION MODEL FOR ANTI-PHISHING USING GENERATIVE AI

Abstract

Phishing attacks are one of the most common cybersecurity threats targeting individuals and organizations through deceptive emails, websites, and messages. Traditional phishing detection systems rely heavily on static datasets and predefined rules, which often fail to adapt to newly evolving phishing techniques. This research proposes an Automated Scenario Generation Model for Anti-Phishing using Generative AI that dynamically creates realistic phishing and legitimate communication scenarios to enhance detection and training systems. The proposed model utilizes Generative Artificial Intelligence techniques such as Generative Adversarial Networks (GANs) and Large Language Models (LLMs) to simulate diverse phishing patterns including email spoofing, malicious links, and social engineering messages. These generated scenarios are used to train machine learning classifiers for improved detection accuracy. The system also evaluates generated scenarios using linguistic, structural, and behavioral features to distinguish phishing attempts from legitimate communications. Experimental results demonstrate that the generative approach improves phishing detection performance and enhances cybersecurity awareness by providing dynamic training datasets. This model contributes to proactive cybersecurity defense by continuously adapting to emerging phishing strategies.