AI-Powered Cybersecurity: How machine Learning is Automating Threat Mitigation

Main Article Content

Mohit Raikwar, Shantilal Bhayal, Digendra Singh Rathore, Shweta Gupta

Abstract

The fast tracking of cyber threats has necessitated for practical and adaptable solutions to protect digital infrastructure in the fast pace,. In this work, we demonstrate practical implementation of an AI powered cybersecurity framework, which takes the form of an ML based framework that automates threat detection and mitigation. Supervised learning is used to detect patterned attacks, while unsupervised learning learns novel threats in real world environments from static, rule based systems. The framework was then deployed at a production tested level in an enterprise grade, simulating live cyber attack scenarios (including phishing, ransomware, and insider attack). Real time threat identification combined with automated containment and remediation protocols were realized by integrating ML models into an intrusion detection system (IDS) pipeline. As they demonstrated a 90 percent reduction in response times and a 30 percent improved detection of zero day attacks than traditional methods. Finally, it finds that AI driven solutions are viable and seem applicable to automate tasks out of human work in operational security and preemption of future threats. In other words, it is a roadmap for organizations which wish to move from reactive to intelligent, responsive cybersecurity approaches.

Article Details

Section
Articles