AI-Driven Cloud Malware Detection and Classification System

Abstract

The growing number of cyber threats, especially in cloud systems over the last few years has increased the demand for speedy and reliable incident response schemes. This project has implemented an AI-based incident response system to better detect and respond to security incidents in cloud environments. You use machine learning methods with cloud platforms such as the Google Cloud and Microsoft Azure with better efficiency and scalability to your work. Flask is used to develop an automated pipeline, composed of modules for network traffic classification, web intrusion detection and incident-based malware analysis. The system was evaluated by using publicly available datasets such as NSL-KDD, UNSW NB15, and CICIDS-2017. As a result, the Random Forest algorithm yielded accuracies of 90%, 75% and 90% for these datasets and moreover also had a precision rate as high as 96% in malware analysis. And also a neural network model then we got an accuracy of 90%. To manage the high processing requirements, cloud-based GPUs and TPUs are utilized, ensuring faster computation. Containerization techniques are also applied to make the system flexible, scalable, and easy to deploy across different cloud platforms. Overall, the proposed system helps in reducing the time required to respond to security incidents, lowers operational risks, and provides a costeffective solution. This project demonstrates how combining artificial intelligence with cloud infrastructure can improve modern cyber security practices.