Proposing a New Foundation of Attack Trees in Monoidal Categories

This short paper introduces a new project studying at the intersection of threat analysis using attack trees and interactive theorem proving using linear logic. The project proposes a new semantics of attack trees in dialectica spaces, a well-known model of intuitionistic linear logic, which offers two new branching operators to attack trees. Then by exploiting the Curry-Howard-Lambek correspondence it seeks to develop a domain-specific linear functional programming language called Lina – for Linear Threat Analysis – for specifying and reasoning about attack trees.