Administrative liability for the unlawful use of artificial intelligence

Abstract

The world has witnessed rapid development in the field of artificial intelligence, thanks to the increase in data in information systems. Artificial intelligence is the theory of the development of machines to enable them to perform tasks that normally require human intelligence, such as the ability to think, perceive, and solve problems by learning from mistakes and previous experiences to gain expertise, and to unify decisions through the method of collecting and analyzing data and information, for speech recognition and language translation. Moreover, it has become involved in legal relationships and concludes legal transactions, providing opportunities for improved performance, speed, and accuracy. However, despite the existence of a significant gap between current legislation and AI classifications, especially regarding information security and digital trust, which complicates the process of regulating it. Simulations coupled with individual behavior, such as deep fakes, may result in harm to others that is difficult to address under traditional liability rules. This is due to the autonomous decision-making process, without receiving any orders from the person who created the simulation. This makes it difficult to control. This has prompted some researchers to consider recognizing legal personality. The purpose of recognizing legal personality is not to establish the capacity of artificial intelligence to exercise rights, but rather to determine the person responsible for the harm. Although technology can provide services to humans at a speed exceeding that of humans, it may still cause harm that makes it difficult to determine who is responsible for the harm. The importance of this study lies in identifying the basis for administrative liability for errors in the administration's use of artificial intelligence technologies in administrative decision-making.