Risk Evaluation Methodology for AI-enabled Cybersecurity  in Federated Mission Networks

Esteves,Eduardo José da Costa Pinto Silva

http://hdl.handle.net/10400.26/59797

Use this identifier to reference this record.

Name:	Description:	Size:	Format:
Dissertação_de_Mestrado_AlfAl_441_Eduardo_Esteves_AM_Finalizado.pdf		435.49 KB	Adobe PDF	Download

Send Feedback

Authors

Esteves,Eduardo José da Costa Pinto Silva

Advisor(s)

Correia, Miguel Nuno Dias Alves Pupo

Abstract(s)

The increasing complexity of cyber threats necessitates the development of more advanced cybersecurity tools. This thesis introduces an Artificial Intelligence (AI) risk evalua tion methodology for AI-enabled cybersecurity solutions, focusing on Federated Mission Networks (FMNs), a crucial component in military and governmental operations. The methodology is based on established methodologies such as OWASP’s risk rating method ology and the STRIDE threat model to systematically assess the risks AI introduces in these environments. The methodology evaluates AI-driven tools based on their ability to detect, mitigate, and respond to cyber threats while identifying potential vulnerabilities, such as spoofing, tampering, and denial of service. The core of this research focuses on ensuring that AI systems are not only practical but also secure, emphasizing the importance of continuous monitoring, robust authentication mechanisms, and real-time threat detection in mission-critical networks. The methodology is applied to tools like OutGene and Cortex XDR, amongst others, making a total of 9 tools, illustrating the need for rigorous AI evaluation in safeguarding sensitive infrastructures. By addressing AI-specific vulnerabilities and aligning security practices with the unique requirements of FMNs, this thesis contributes to the growing f ield of AI-enhanced cybersecurity, offering a structured, scalable approach to ensuring network resilience and operational integrity.