Thesis topic:

An Ontology-Based Mapping and Evaluation of Security Risk Management for AI-powered Systems

Supervisor: Ijeoma Faustina Ekeh
Co-Supervisor: Mubashar Iqbal, PhD
contact: ijeomafa@ut.ee
This thesis seeks to advance and develop a comprehensive ontological tool for threat modelling of AI/ML systems. The goal is to create a structured framework using an already established conceptual framework (Threat Modelling for AI/ML Systems) that captures and organises these concepts related to threats, vulnerabilities, assets, and mitigations and the lifecycle specific to AI/ML. The ontology should help formalise the relationships between various components in AI/ML systems and their associated risks, and evaluate the effectiveness of this tool.