Main Article Content
Phishing attacks continue to pose significant risks to individuals and organizations, emphasizing the need for effective detection mechanisms. This work proposes a novel approach that combines Generative Adversarial Networks (GANs) with machine learning (ML) techniques, specifically the Random Forest Classifier (RFC), to enhance phishing detection capabilities. The GAN component generates synthetic phishing examples that closely resemble real-world attacks, augmenting the training data and improving the model's ability to generalize. The ML component, leveraging these synthetic examples, analyzes relevant features and patterns to accurately classify phishing websites. The proposed system offers several advantages, including enhanced detection accuracy, adaptability to evolving threats, improved generalization, discovery of discriminative features, and flexibility in selecting and integrating ML algorithms such as RFC. Experimental evaluation demonstrates the effectiveness of the combined GAN-ML-RFC approach, providing a promising solution for robust and adaptive phishing detection systems.