Open AccessToward to Information Security of AI-Enhanced Weapons
Author(s) -
Vadim Gribunin,
Sergey Kondakov
Publication year - 2021
Publication title -
voprosy kiberbezopasnosti
Language(s) - English
Resource type - Journals
ISSN - 2311-3456
DOI - 10.21681/2311-3456-2021-5-5-11
Subject(s) - computer security , software deployment , computer science , information security , field (mathematics) , information security standards , security information and event management , risk analysis (engineering) , security service , cloud computing security , software engineering , network security policy , cloud computing , mathematics , pure mathematics , operating system , medicine
Purpose of the article: Analysis of intellectualized weapons using machine learning from the point of view of information security. Development of proposals for the deployment of work in the field of information security in similar products. Research method: System analysis of machine learning systems as objects of protection. Determination on the basis of the analysis of rational priority directions for improving these systems in terms of ensuring information security. Obtained result: New threats to information security arising from the use of weapons and military equipment with elements of artificial intelligence are presented. Machine learning systems are considered by the authors as an object of protection, which made it possible to determine the protected assets of such systems, their vulnerabilities, threats and possible attacks on them. The article analyzes the measures to neutralize the identified threats based on the taxonomy proposed by the US National Institute of Standards and Technology. The insufficiency of the existing regulatory methodological framework in the field of information protection to ensure the security of machine learning systems has been determined. An approach is proposed that should be used in the development and security assessment of systems using machine learning. Proposals for the deployment of work in the field of ensuring the security of intelligent weapons using machine learning technologies are presented.