AI-Driven Threat Detection : Enhancing Security and Reducing Human Intervention

Abstract

The need for artificial intelligence in cyber security frameworks has grown significantly with the evolution of cyber threats. Traditional rule-based methods, especially in areas like threat detection and log analysis, are struggling to keep up with the ever-growing volume and complexity of data generated by modern systems. As attackers adopt more advanced and evasive techniques, there is a clear need for intelligent systems that can adapt and respond in real time. AI-driven methods, leveraging machine learning, offer significantly more efficient, accurate and reliable solutions to these problems. These methods can process vast amounts of data with speed and precision, detect subtle anomalies that may escape human analysts, and continuously improve over time as they are exposed to new threats.

This thesis was conducted mostly as a literature review, with the goal of compiling a thorough analysis of AI-based threat detection techniques and potential improvements to certain security systems. The research examines a range of supervised, unsupervised, semi-supervised, reinforcement and deep learning approaches, discussing their strengths, limitations and practical applications within cybersecurity environments. In doing so, it not only incorporates current industry findings, but also identifies gaps and emerging trends in the field. By reviewing existing research and real-world implementations, this thesis highlights how cyber security systems can be enhanced by the help of AI. Ultimately, the study aims to provide valuable insights for researchers seeking to develop more adaptive and resilient threats detection systems in the face of increasingly complex cyber threats.