Revisiting Attack Trees for Modeling Machine Pwning in Training Environments

In this paper, we propose a machine pwning modeling that helps as an easy-to-follow guide for ethical hackers, whether beginners or professionals, in training environments. In particular, we build a mini dataset of attack trees. Each attack tree (or tree of hints) connects multiple types of blocks representing a series of ethical hacking actions in a formalized flowchart-like format. More precisely, we modeled machine pwning using a unified attack tree format for popular vulnerable machines typically used for ethical hacking training. The paper presents a human study for usability, where we found that our mini dataset helps reduce the time to solve (TTS) in most scenarios. In findings, an attack tree is a quick and concise cheat sheet replacing a long and detailed walkthrough, can be gamified, and, therefore, is excellent training material for trainees with diverse technical backgrounds and expertise. We present preliminary ideas for employing these trees by artificial intelligence.