An Artificial Intelligence Approach for Deploying Zero Trust Architecture (ZTA)

Abstract

Cybersecurity is critical in preventing infractions, maintaining digital workplace discipline, and ensuring that laws and regulations are obeyed. Zero Trust Architecture (ZTA), often known as perimeter-less security, is a novel method for designing and implementing secured IT systems. Zero trust's basic notion is “never trust, always verify,” which indicates that devices should not be trusted by default. This means that each access from or to any asset must be assessed and follow the standard guidelines of the organization. Maintaining this type of control imposes a high burden on IT security and system administrators to be able to track and validate each control and manually sustain the configuration needed. With the power of Classification Algorithms in Machine Learning, we will explore in this paper an alternative solution to save time and effort and help maintain the same security posture with less human intervention. The proposed approach utilizes the information from available security feeds and statically configured policies to enforce and maintain zero-trust network policies. By analyzing the data, it will be feasible to identify the required policies to be configured and compare them against the traditional compliance rules to auto-configure the policies. This approach aims to enhance the existing security intelligence engines with more sophisticated rules and less time and effort.