The World Wide Web offers vast information and opportunities but also harbors malicious websites designed to steal data and launch attacks. Current methods to combat these threats analyze downloaded webpages, detecting dangers only after the damage is done. This research proposes a proactive solution: a system that analyzes URLs themselves, identifying threats before users click on risky links, like a security guard spotting suspicious activity. Traditional URL analysis methods focus on specific features and are easily outsmarted by evolving attack tactics. This study employs sophisticated algorithms like decision trees and Naive Bayes to examine every detail of a URL—its structure, domain name, and hidden clues—to uncover malicious intent. The algorithms are trained using data from known phishing sites and legitimate website directories, learning patterns to identify future threats more accurately. The research highlights the need for innovation in URL security. As cybercriminals develop new tactics, our defenses must evolve. Implementing proactive URL analysis can create a safer web experience, preventing users from entering the web's dark alleys and falling victim to malicious attacks. This approach ensures a more secure and trustworthy online environment by staying ahead of cybercriminals with dynamic and evolving defenses.