인공지능의 법적정의와 인공지능기술 발달에 따른 국가의 기본권 보호의무

Abstract

Recently, the development of research on deep learning, a type of machine learning, has rapidly developed artificial intelligence. There are many examples of human-made machines exceeding human abilities, but artificial intelligence differs from other machines in that it is a entity with intelligence, which is an essential ability of humans. These differences bring not only expectations but also concerns about artificial intelligence, which provides the need for constitutional discussions on artificial intelligence. The fulfillment of the state’s duty to protect fundamental rights generally results in limiting the basic rights of other individuals. Laws to protect individuals who may be infringed on basic rights by the use of artificial intelligence limit the freedom of occupation of those who want to develop and use artificial intelligence. Laws that limit basic rights must meet the principle of clarity, so it is required that the subject of discipline be clearly defined. Therefore, the artificial intelligence law must first clearly define the concept of artificial intelligence. If related laws are enacted without artificial intelligence being properly defined, there is a risk of violating individual freedom of occupation while failing to properly fulfill the obligation to protect basic rights. Methods of approaching artificial intelligence may include focusing on human beings and electronically implementing their abilities, and focusing on creating an electronic entity that can make the most rational decisions under uncertain circumstances. However, since most of the current artificial intelligence research is centered on the latter, and it is difficult to see that the reality of artificial intelligence currently used embodies human intellectual ability, the latter should be focused when defining the concept. It would be reasonable to assume that it is the realm of biology, not the realm of computer science using electronic methods. If so, I think it is correct to call the legal concept of artificial intelligence a “software or hardware system designed by humans, which recognizes the environment through data analysis under complex environments and selects and executes actions that best suit the achievement of a given goal with information and knowledge.” Of course, this definition may not be clear, but artificial intelligence is a continuously developing field and its application fields vary, so it is necessary to leave the concept open to some extent. If the Artificial Intelligence Act is approached excessively from the perspective of regulation, it can block the development of artificial intelligence, which is not the right phenomenon in light of the economic value that artificial intelligence can create. For now, I think it is a desirable direction to create an organization that can conduct democratic and professional deliberations rather than to contain a lot of content in the law so that various dispositions can be made. It may be considered premature to establish as a central administrative agency, and if so, I think it is one way to establish a committee in the Ministry of Science and ICT and recognize its binding force on the resolution.