Restrictions and liability for the use of artificial intelligence in armed conflicts: national and international dimensions
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
The article examines existing international and national approaches to restricting artificial intelligence used in the military industry. It questions whether the Geneva Conventions provide a sufficient number of safeguards for subjects of international humanitarian law who intend to use artificial intelligence during armed conflicts. The historical background of the concept of «artificial intelligence» is considered on the example of national theories of scientists. Parallels are drawn between the legal personality of an individual and a computer endowed with mental abilities, thinking, and human behavior in order to clarify the possibility of establishing liability for artificial intelligence. It has been established that the «artificiality» of intelligence in military systems is not autonomy in the sense of human autonomy of will. And the legal personality of artificial intelligence excludes the possibility of holding a computer liable due to the lack of a legal vision of artificial intelligence as a full-fledged subject of legal relations. The issue of liability for the use of artificial intelligence in armed conflicts within the framework of international subjects – states and international organizations, as well as national ones – commanders, manufacturers, developers, and operators, is considered in detail. Based on the results of the research, it is concluded that restrictions on the use of artificial intelligence should be reduced to a direct ban on algorithmic thinking in modern weapons without direct human control. In order to speak about international legal liability for violation of the laws of war, the combatant must remain the only subject of prosecution. And artificial intelligence in weapon systems must remain a non-autonomous source of information for battlefield analysis by an operator or commander. Such as the use of artificial intelligence in air defense systems, satellites, or reconnaissance drones. Any lethal weapon involving artificial intelligence is limited by the «human-in-the-loop» model, and the adoption by states of the Geneva Conventions and their Additional Protocols established safeguards regarding the use of autonomous weapons during armed conflicts.
Fields of Study & Subjects
International Law and Human Rights
Digitalization
Law
and Regulation
Military Strategy and Technology
Publication details
- Journal / source
- Uzhhorod National University Herald Series Law
- Volume
- 5
- Issue
- 93
- Pages
- 361–366
- Publication date
- 31 March 2026
- Language
- EN