PhishGuard AI: An Intelligent System for Phishing Detection and User Education using Explainable AI

In today’s fast-paced digital environment, phishing remains a persistent and damaging threat. Over 90% of successful cyberattacks originate from a “Human Error Gap,” such as clicking a deceptive link. To address this, PhishGuard AI has been developed—an AI-powered cybersecurity system designed to combat phishing through real-time detection and transparent user education. This intelligent system is capable of analyzing complex URL features to predict malicious intent with high accuracy (over 93%). Built using an optimized XGBoost Classifier for backend analysis and a React.js browser extension for an intuitive front-end interface, the system’s core innovation is its Explainable AI (XAI) module. This module translates complex predictive scores into clear, human-readable risk factors (e.g., “The URL contains the ‘@’ symbol”). This transparency serves as a real-time educational tool, empowering the user to understand threats and build digital resilience. This project explores the design, development, and deployment of this system, emphasizing the mitigation of human error by transforming security from a “black box” into a transparent educational partner.