Complex cyber-physical systems (CPSs), such as AI-enabled CPSs, integrate various technologies leading to an expansion of the attack surface over time. Traditional mitigation methods struggle to keep pace with evolving adversaries’ strategies. In this context, AI can offer powerful opportunities for dynamic and proactive threat prediction, enabling the anticipation of risks before they actually occur and adapt defenses accordingly.
Main goal: This proposed master thesis aims to explore dynamic AI approaches for predicting potential malicious behaviour of AI-enabled cyber-physical systems.
References:
Apruzzese, Giovanni, et al. "The role of machine learning in cybersecurity." Digital Threats: Research and Practice 4.1 (2023): 1-38.
Recommended: 60 credits (but can be adapted to 30 credits if needed)