Year: 2026
Type: article
Abstract: Phishing attacks have become very common these days, and many users find it hard to identify whether a link or message is genuine or not. Attackers keep changing their tricks, which makes traditional detection methods less effective. To deal with this problem, an AI-based real-time phishing detection system was developed that can check suspicious URLs and messages and classify them as safe, suspicious, or phishing. The system works by looking at different patterns such as strange domain names, use of special characters, fake login or verification messages, shortened links, and content that creates urgency to mislead users. Instead of depending only on predefined rules or blacklists, it uses a combination of AI analysis and simple rule checks so that the results are more reliable and practical. The whole system is designed in a way that users can easily scan links either through a web interface or directly using a Chrome extension, while the backend processes the input and returns the result along with a confidence score and explanation. After testing with different kinds of inputs, it was observed that the system can correctly detect most phishing attempts, even when the links are new or slightly modified. It also responds quickly, which makes it suitable for real-time usage. Compared to older methods, this approach shows better performance in identifying unknown threats. Overall, the work shows that combining AI with simple detection logic can help in building a useful tool for improving online safety, and it can be further used in areas like secure browsing, email filtering, and fraud prevention.
Language: English
Topic: Spam and Phishing Detection
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7154402076
FWCI:
Cites:
Cited by:
Related to: