Machine Learning Applications for Insider Threat Detection in Cloud Security: A Narrative Review

Authors: Oladimeji Ganiyu Lawal Olufunmilayo Oyelowo Rofiat(Computer Science Dept. Moshood Abiola Polytechnic, Abeokuta)

Journal: Nigerian Journal Of Technical Education | Volume: 24 | Issue: 1 | Pages: 92 - 98

Keywords:Machine Learning; Cloud Security; Insider Threats; Anomaly Detection; User Behavior Analytics; Privacy.

Abstract

This narrative review aims at synthesizing recent literature that explores machine learning (ML) for addressing insider threats in cloud security environments. In 2023, a systematic search of the literature was performed for peer-reviewed papers, conference proceedings, and industry reports related to ML techniques, challenges, and real-world problems from 2018 onwards. Some of the supervised and unsupervised learning techniques used to detect malicious activities in this review are Radial Basis Function Neural Networks and Random Forests. We offer a new comparison of these techniques, examining how well each performs in various insider threat scenarios. ML improves threat detection, but challenges remain with data accessibility, feature selection, and ethical issues. The research demonstrates the need for a balanced approach leveraging both ML and human-centric strategies to achieve effective insider threat mitigation. Finally, we discuss future research directions, focusing on building transparency into AI-driven systems and exploring federated learning techniques to overcome current limitations.