인공지능(AI)과 관련한 법률관계

Legal relations related to Artificial Intelligence

When economic value is created by an Artificial Intelligence, the acquisition of that value is generally attributed to the owner of the AI. At the same time, even if damage is caused to others due to autonomous judgment or malfunction of the AI, civil liability is generally imposed on the owner. However, even in cases where the AI commits legal or illegal acts beyond human control, will the effects be attributable to the AI itself and will the legal responsibility can be imposed on the AI itself? It is not clear whether it is. In that case, although researchers are discussing how to control the interests of the other party who engages in legal acts with AI or is harmed by illegal acts by AI, no concrete legal or institutional mechanism has been established. Since the degree of evolution of AI is still within the control of humans, these discussions have not yet begun in earnest, but considering the speed of its development, this problem will become a reality in the near future. If social necessity and social value judgment are premised, it will be possible to proceed with legislative procedures to grant the legal personhood to the AI, which is a new entity. However, even if the AI has the legal personhood, the scope of it should be limited, as seen in corporations. Since the scope of the legal personhood of the AI is the one of the owner’s need or convenience, it is reasonable to ensure that the scope is specified in the AI register upon the application of the owner, and to grant the legal personhood only within that scope. In addition, it is appropriate to allow the AI to be granted legal personhood on the condition that it maintains registered minimum liability property, and to deny legal personhood in cases where the minimum liability property cannot be maintained due to the actions of AI. In this way, by setting the minimum liability property and ensuring that the specified property is maintained, it is possible to protect the other party transacting with the AI or the victims who have suffered damage due to the actions of AI.