Poster: Formalizing Cognitive Biases for Cybersecurity Defenses

Abstract

As network attacks are becoming increasingly sophisticated, there is a need to develop new defense techniques. Recent work has demonstrated the successful use of cognitive biases to obstruct progress in network attacks by constructing scenarios which exploit biases in attackers. While informal definitions of many cognitive biases are well-established, the lack of definitions in a formal language limits the ability to construct scenarios in an automated fashion. This work proposes formal definitions using first-order logic to encapsulate the complex relationships between actions and biases.