Abstract
Phishing attacks are an increasingly common cyber threat that exploit the trust users place in genuine websites by mimicking their look and feel. This research proposes a novel AI-based detection system designed to identify phishing domains that imitate authentic websites. Our approach leverages machine learning algorithms to analyze visual similarities, domain patterns, and metadata between phishing and legitimate sites. Through this method, the system detects phishing attempts before users are deceived. Key results indicate a high detection accuracy and a reduction in false positives, which demonstrate the potential of this method for enhancing cybersecurity. Future improvements in realtime detection and integration into security protocols are discussed.