An Ontology-Based Cybersecurity Framework for AI-Enabled Systems and Applications

Ontologies have the potential to play an important role in the cybersecurity landscape as they are able to provide a structured and standardized way to semantically represent and organize knowledge about a domain of interest. They help in unambiguously modeling the complex relationships between various cybersecurity concepts and properties. Leveraging this knowledge, they provide a foundation for designing more intelligent and adaptive cybersecurity systems. In this work, we propose an ontology-based cybersecurity framework that extends well-known cybersecurity ontologies to specifically model and manage threats imposed on applications, systems, and services that rely on artificial intelligence (AI). More specifically, our efforts focus on documenting prevalent machine learning (ML) threats and countermeasures, including the mechanisms by which emerging attacks circumvent existing defenses as well as the arms race between them. In the ever-expanding AI threat landscape, the goal of this work is to systematically formalize a body of knowledge intended to complement existing taxonomies and threat-modeling approaches of applications empowered by AI and to facilitate their automated assessment by leveraging enhanced reasoning capabilities.