사상의 자유의 재조명 ― 인공지능에 의한 의사형성과정의 조종 가능성을 중심으로 ―

The Constitution of Korea provides for the protection of people's internal intellectual activities by guaranteeing various rights to intellectual freedom, such as freedom of conscience, religion and thought. So far, discussions on the protection of intellectual activities have been based only on state interference where the results of intellectual activities are expressed or materialised externally. With the development of artificial intelligence, however, it is possible to analyse inner sanctum of human beings and to influence the decision-making process. It is possible to influence the decision-making process through the analysis of accumulated data on human behaviour. In light of the increasing need to protect decision-making processes, this paper will delve into freedom of thought, which has been relatively unexplored. The development of artificial intelligence has reached a stage where it is possible to intervene in an individual's decision-making process, inducing or distorting behaviour or expression that would not have occurred or expressed without such intervention. This article explains the danger of AI through the concept of ‘manipulation’ and examines the constitutional issues that can arise when AI manipulates political decision-making. The paper goes on to argue that the obligation of government to protect fundamental rights must be extended to intellectual fundamental rights, including freedom of thought, to address issues of democratic legitimacy, the outcome of elections that might be manipulated by AI, and the emergence of private entities using AI to infringe fundamental rights. In order to meet the obligation to protect fundamental rights, it is necessary to establish systematic rules to protect people from the risks posed by artificial intelligence, and confirms that manipulation by artificial intelligence can constitute a serious violation of fundamental rights through the content of currently proposed legislation.