Attack Tree Generation by Policy Invalidation

Attacks on systems and organisations increasingly exploit human actors, for example through social engineering, complicating their formal treatment and automatic identification. Formalisation of human behaviour is difficult at best, and attacks on socio-technical systems are still mostly identified through brainstorming of experts. In this work we formalize attack tree generation including human factors; based on recent advances in system models we develop a technique to identify possible attacks analytically, including technical and human factors. Our systematic attack generation is based on invalidating policies in the system model by identifying possible sequences of actions that lead to an attack. The generated attacks are precise enough to illustrate the threat, and they are general enough to hide the details of individual steps.