DEVELOPMENT OF A CYBERSECURITY MODEL FOR ARTIFICIAL INTELLIGENCE SYSTEMS BASED ON A MULTI-LEVEL PROTECTION ARCHITECTURE AND INTELLIGENT RISK ASSESSMENT
True Masters Innovations · 2026
DOI: 10.65769/8guubm2wws ↗Kurz gefasst
Fachartikel in der Zeitschrift True Masters Innovations, 2026. Verfasst von Максат Серикулы Мубарак und 1 weiteren Autor:innen. Die Arbeit nennt 5 Literaturangaben.
Abstract (Originalsprache)
Annotation. The rapid development of artificial intelligence (AI) technologies and their widespread adoption in public administration, industry, healthcare, education, the financial sector and the defense sector are accompanied by the emergence of new threats to information security. Modern artificial intelligence systems, including large language models, generative models, and intelligent decision support systems, have specific vulnerabilities that are not taken into account by traditional cyber defenses. The most common threats are attacks on training data, extraction of model parameters, implementation of adversarial examples, circumvention of security mechanisms through Prompt Injection and Jailbreak, as well as leakage of confidential information. The article analyzes modern approaches to ensuring cybersecurity of artificial intelligence systems, examines international standards and recommendations in the field of AI Security, including the NIST AI Risk Management Framework, MITRE ATLAS and OWASP Top 10 for Large Language Model Applications. Based on the analysis of existing research, the limitations of existing protection methods have been identified, which consist in the lack of an integrated approach to ensuring security at all stages of the life cycle of an intelligent system. The author's multilevel model of ensuring cybersecurity of artificial intelligence systems is proposed, including levels of data protection, machine learning models, user interaction, computing infrastructure, security monitoring and risk management. Unlike existing solutions, the model integrates classical information security mechanisms with intelligent threat analysis tools and adaptive risk assessment, which makes it possible to increase the resilience of AI systems to modern cyber attacks. The practical significance of the research lies in the possibility of applying the proposed model in the development of secure intelligent information systems operating in government, corporate and military organizations.
Zitieren
APA
Мубарак, М. С., & Түлкібай, Д. Ә. (2026). DEVELOPMENT OF A CYBERSECURITY MODEL FOR ARTIFICIAL INTELLIGENCE SYSTEMS BASED ON A MULTI-LEVEL PROTECTION ARCHITECTURE AND INTELLIGENT RISK ASSESSMENT. True Masters Innovations. https://doi.org/10.65769/8guubm2wws
Vancouver
Мубарак МС, Түлкібай ДӘ. DEVELOPMENT OF A CYBERSECURITY MODEL FOR ARTIFICIAL INTELLIGENCE SYSTEMS BASED ON A MULTI-LEVEL PROTECTION ARCHITECTURE AND INTELLIGENT RISK ASSESSMENT. True Masters Innovations. 2026. doi:10.65769/8guubm2wws
BibTeX
@article{2026,
author = {Мубарак, Максат Серикулы and Түлкібай, Дінмұхамед Әбділдаұлы},
title = {DEVELOPMENT OF A CYBERSECURITY MODEL FOR ARTIFICIAL INTELLIGENCE SYSTEMS BASED ON A MULTI-LEVEL PROTECTION ARCHITECTURE AND INTELLIGENT RISK ASSESSMENT},
journal = {True Masters Innovations},
year = {2026},
doi = {10.65769/8guubm2wws}
}
EUVIMED