Building a Use Case-Driven Adversarial Threat Matrix: A Comprehensive Approach

Prodromos Kiriakidis, Efi Kafali, Georgios Rizos, Antonis Voulgaridis, Theodoros Semertzidis, Stephane Gazut, Petros Daras, Konstantinos Votis

Abstract: To effectively safeguard against increasingly so-phisticated and diverse cybersecurity threats, particularly those driven by Artificial Intelligence, a comprehensive understanding of the threat landscape is essential. A multifaceted threat matrix is introduced in this paper, developed within the KINAITI CS project and derived from the analysis of six diverse use cases and insights gathered from expert workshops. The matrix aims to highlight and prioritize emerging threats posed by AI -driven attacks and vulnerabilities.