Scientific and Security Research
Automated Penetration Testing Based on Agentic AI
Development of a dynamic system based on Large Language Models (LLMs) for goal-oriented guidance and automation of network and infrastructure security assessment processes.

01 . INTRODUCTION
Necessity and Challenges of Traditional Penetration Testing
With the increasing complexity of IT infrastructures and the ever-expanding cyber threats, Penetration Testing has become one of the most important and critical security assessment methods for organizations. However, relying solely on traditional methods does not meet modern needs.