AI-BASED EMPOWERING OF COMPLIANCE WITH ZERO TRUST PRINCIPLES IN CORPORATE NETWORK ARCHITECTURE

Abstract

Modern corporate network infrastructures are characterized by a continuous increase in the number of information services, the widespread use of hybrid and cloud environments, and the growing complexity of network access policies. The dynamic nature of such environments leads to frequent changes in network security configurations, while requirements governing interactions between corporate services may be maintained and updated across different information systems independently of the actual network configuration. As a result, discrepancies may gradually emerge between declared requirements and the network access rights actually granted, making it more difficult to maintain an appropriate level of security and compliance with the principles of Zero Trust Architecture.

This problem is particularly relevant in large corporate environments, where the number of network access rules may reach several thousand, while their periodic analysis and access rights recertification are time-consuming processes that require substantial human and operational resources. Traditional configuration analysis tools can identify certain anomalies; however, they are not always capable of determining the extent to which the access rights actually granted correspond to their original purpose and to the current security requirements of corporate services. This creates conditions for the accumulation of outdated, redundant, undocumented, or overly permissive access rights and complicates the implementation of the principles of least privilege, explicit authorization, and continuous verification of security policies.

This paper investigates the potential use of artificial intelligence technologies to improve the efficiency of network access policy control and recertification processes in the context of implementing Zero Trust principles. A conceptual approach is proposed that focuses on the automated processing of heterogeneous information about corporate services and the actual state of network security policies in order to assess their compliance with security requirements and Zero Trust principles.

The proposed approach involves the use of intelligent configuration analysis to identify potential deviations, detect network access rules that require additional review, and generate recommendations for responsible specialists. In this context, artificial intelligence is considered a supporting mechanism for decision-making rather than an autonomous tool for modifying security policies. The application of this approach is expected to reduce the amount of manual work required during recertification, improve the scalability of the control process, and support the continued alignment of network policies with Zero Trust principles in complex corporate infrastructures.