AI-Powered Cyber Threat Intelligence Platform

Abstract

Cybersecurity threats are becoming more complex and challenging to identify using conventional security measures. While traditional signature-based systems work well against known assaults, they frequently miss new threats like ransomware, zero-day exploits, and advanced persistent threats. This study suggests an AI-Powered Cyber Threat Intelligence (CTI) Platform that combines behavioral analysis and machine learning methods for proactive threat identification and categorization in order to address these issues. The platform uses a Long Short-Term Memory (LSTM) model to identify anomalies in sequential log data and a Random Forest Classifier to classify network traffic. Data is gathered from publicly accessible cybersecurity sources, system logs, and network traffic databases. Data collection, preprocessing, model building, threat detection, and dashboard visualization layers make up the suggested design. A classification accuracy of roughly 97.8% is shown by experimental evaluation using the CICIDS2017 dataset, with precision and recall levels surpassing 96%. The findings show that the suggested platform reduces the need for human monitoring while efficiently identifying cyber threats. For academic institutions, researchers, and small businesses looking to improve their cybersecurity 
capabilities, the platform provides a scalable and affordable alternative.