Phishing attacks have become one of the most serious cybersecurity threats worldwide, causing identity theft, financial loss, and data breaches. Attackers use fake websites, emails, and malicious links to trick users into revealing sensitive information. Traditional security mechanisms such as antivirus software and browser filters are often unable to detect newly generated phishing URLs, making users vulnerable to attacks. To address this issue, this project proposes a Real-Time Automatic Phishing Detection System that identifies and blocks phishing links instantly. The system uses Machine Learning techniques, specifically the Random Forest Classifier, to analyze URL features such as length, domain age, and special characters. It operates silently in the background without requiring user intervention, ensuring continuous and seamless protection. The system is developed using Python, Java, JavaScript, Node.js, MongoDB, HTML, and CSS to support multi-platform functionality. It provides real-time alerts and maintains logs of detected threats for further analysis. The proposed solution aims to enhance cybersecurity by offering proactive protection and ensuring a safer digital environment for individuals and organizations.