AI-Powered Penetration Testing using Shennina: From Simulation to Validation
Stylianos Karagiannis, Camilla Fusco, Leonidas Agathos, Wissam Mallouli, Valentina Casola, Christoforos Ntantogian, Emmanouil Magkos · 2024
Artificial intelligence has been greatly improved nowadays, providing innovative approaches in cybersecurity both on offensive and defensive tactics. AI can be specifically utilized to automate and conduct penetration testing, a task that is usually time-intensive, involves high-costs, and requires cybersecurity professionals of high expertise. In this research paper, we utilize an AI penetration testing framework to validate, discover and analyze the techniques that were used. To this end, we conducted a validation process in a realistic environment and to collect the relevant datasets from the execution of the cyberattacks. Finally, the behavior of the AI penetration testing was analyzed in order to adapt and upgrade further. Overall, the research paper provides contributions to dataset generation and a methodology to understand the details of the attack simulation.