Cybersecurity Challenges of the Integration of Artificial Intelligence (AI) Solutions
Hadmérnök · 2025 · 20 (4) · 17-38
DOI: 10.32567/hm.2025.4.2 ↗Kurz gefasst
Fachartikel in der Zeitschrift Hadmérnök, 2025. Verfasst von Károly Kassai. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
The applicability of artificial intelligence (AI) is no longer in question today. AI has become so integrated into countless areas of the economy and society that we no longer even notice that a given service is provided by an AI system. The use of AI systems is a sensitive issue in many areas, such as the armed forces. This paper draws attention to the diversity of military applications of AI and the general risks involved. The armed forces need to use not just one or two AI systems, but dozens of AI solutions that must be integrated into military information systems. These AI systems perform tasks at different operational levels or support the operation of other systems. The study provides guidance on the most important steps in the necessary risk management, based on the legal framework, standards and best practices. Tailored risk management provides the basis for local, system-specific regulation of military organisations, which must be compiled from existing cybersecurity framework elements. The study emphasises that AI systems cannot be exempt from cybersecurity regulations, so it is necessary to review and supplement existing tools and provide training.
Zitieren
APA
Kassai, K. (2025). Cybersecurity Challenges of the Integration of Artificial Intelligence (AI) Solutions. Hadmérnök, 20(4), 17-38. https://doi.org/10.32567/hm.2025.4.2
Vancouver
Kassai K. Cybersecurity Challenges of the Integration of Artificial Intelligence (AI) Solutions. Hadmérnök. 2025;20(4):17-38. doi:10.32567/hm.2025.4.2
BibTeX
@article{Kassai2025,
author = {Kassai, Károly},
title = {Cybersecurity Challenges of the Integration of Artificial Intelligence (AI) Solutions},
journal = {Hadmérnök},
year = {2025},
volume = {20},
number = {4},
pages = {17--38},
doi = {10.32567/hm.2025.4.2}
}
EUVIMED