CyberGuard - Guarding Against Deceptive Websites

Phishing attacks pose a growing threat to online security, necessitating the development of highly accurate detection tools. Traditional methods reliant on webpage content features and third-party services often result in high false positives. This paper presents CyberGuard, a pioneering phishing detection model that exclusively leverages website URLs, mitigating the need for third-party services. The model harnesses deep learning, employing a Generative Adversarial Network (GAN) framework with LSTM for URL generation and CNN for discrimination.