A Machine Learning Approach to the Evaluation of Zero Trust Compliance in Network Infrastructure

Traditional cybersecurity models wrongly assume that users, devices, and systems inside the perimeter are implicitly trusted. These models have been corrected by assuming no trust with access requests and post-access behaviors. This solution is based on the zero trust (ZT) model, which assumes no implicit trust is granted to assets or user accounts based solely on their physical or network location or asset ownership. ZT has become a popular approach for building secure systems, promoted by industry and government as a new way to produce systems with high security. Security compliance management plays an essential role in mitigating ZT. We propose to evaluate the ZT compliance using a machine learning (ML) approach with random forest. The model will verify the compliance of users, devices, applications, networks, data, and infrastructure in the system. The results show that the model was able to classify the compliance of all six entities with over 95% accuracy. The methodology proposed here can be used as part of a ZT compliance framework in a modern security system.