MIIX GenAI Detect and Analyze Phishing Attack

New Era Institute of Vocational & Continuing Education 43000 Kajang Selangor, DEPARTMENT OF INNOVATION AND TECHNOLOGY Cyber Security (ITCSI)

Supervisor: Ms. Ang Yang Yi

Membership: Chen Yong Xiang, Leong Lap Hang, Lim Jim Hong and Tee Shi Yong

The goal of this project is to combat the concerning increase in digital frauds in Malaysia, with a focus on students, the elderly, and those who are less used to contemporary hazards. With the use of contemporary AI integration, it seeks to create sophisticated tools for phishing detection, malware scanning, and threat analysis to identify and stop frauds. The goal of this project is to prevent online fraud by offering innovative and easily accessible solutions.

The security focus of this project includes real-time incident monitoring, file-based malware scanning, and phishing detection. Users may manually scan files and link scans to certain devices using an easy-to-use online interface. The technology categorizes threat situations according to their level of severity and offers dashboard information. To identify questionable patterns, the threat detection component employs rule-based YARA scanning algorithms.

Because of its scalability, user authentication capabilities, and smooth interface with databases and backend services, Django a powerful Python-based web framework, was chosen to construct the project. This made it perfect for our security application.

Our malware detection method now includes YARA rules, a threat intelligence tool that uses binary pattern matching to categorize and identify malware. This enables us to design unique rules that identify file behaviours or harmful activities.

The system analyses threats and detects phishing attempts using Meta’s LLaMA 3, a large language model. It can detect both overt and covert social engineering techniques since it employs natural language understanding to evaluate email content and spot well-known scam or phishing trends.

source