GAN for Phishing Email Generation and Detection

Our objective for this project is to create a system utilizing Generative Adversarial Networks for the creation and detection of phishing emails to help demonstrate both offensive and defensive measures within cybersecurity. We aim to show the escalating fight between ai-driven threat generation and detection. The generator, or generative model we will use is a LLM to create the realistic emails while the discriminator component that we are envisioning will employ a neural network for the binary classification. We will also create the dynamic structure, combining the two models, for real-time creation and classification. Our timeline includes finding good datasets for the emails, preparing the data, training our two models, integrate the GAN's dynamic structure, test and validate, and prepare the demonstration. This project will hopefully be able to produce a fully-functioning GAN-based system capable of generating and detecting phishing emails to illustrate the effectiveness of AI in understanding and mitigating phishing threats. The interactive portion of the demo would allow people to test how well they are able to differentiate real from fake emails and compare their accuracy to that of the model.

Session Presenter

Joey Cindass

President

Yunseon Choi

Secretary

Olena Molla

Treasurer

The Data Science & AI Club (DS&AI) is a student-led organization that serves as a central hub for students interested in the data sciences and its multifaceted applications in the modern world. We aim to fulfill the needs of ��ɫ�� students by providing opportunities to expand personal and professional networks, and hosting events with an aim to bolster relevant skills applicable to professional development.