인공지능 로봇의 불법행위책임 - 인공지능 로봇 보유자의 과실과 제조판매자의 설명의무 -

Interest in artificial intelligence robots has grown even more during the corona virus. This is because it was thought that the current economic development could be continued while protecting human health and life if production work, transportation work, and medical treatment were carried out through robots that do not get sick or die among the global pandemic. In addition, the commercialization of autonomous vehicles, which has been underway for the past several years, has been sufficient to attract many people's attention, although it has not yet reached full autonomous driving. However, there was an accident that caused human damage during the testing and operation of the autonomous vehicle, and with the advent of weak and strong artificial intelligence robots, the existing liability theory, damage compensation theory, and tort liability were questioned. review was conducted. The theory of holding AI robots accountable is necessary because particularly strong AI will have human-like autonomy. However, there are difficulties in acknowledging act and willful negligence. Moreover, as strong artificial intelligence has not yet appeared, it is only a theoretical discussion. In the midst of this, risk responsibilityism, compensation liabilityism, and benefit liabilityism are argued to overcome the limitations of the existing negligence responsibility. Risk responsibilityism and compensation liabilityism apply existing theories, and benefit liabilityism emerged with the background that the benefits of the invention and development of new technologies should be considered. Realistically, in the current situation, it seems necessary to take an approach with norms that can be used within the current legal system rather than the approach of imposing responsibility or granting legal personality to the AI ​​robot. This is because the legislative solution cannot yet start, and the human and material damage caused by artificial intelligence robots (autonomous vehicles) has already begun. In this regard, it seems reasonable to place a basis for responsibility for the actions of users who initiate the operation of the artificial intelligence robot. Although this seems to impose severe liability on the user, in reality, it facilitates the proof of responsibility based on the product description obligation that the manufacturer/seller or manufacturer/seller has to fulfill to the user, and in many cases, the manufacturer/seller or manufacturer/seller, not the user, is liable. It is an approach to lead Through this, an approach that utilizes the advantages of artificial intelligence robots and protects victims in good faith is expected to be possible. In the case of Germany's autonomous vehicle law (das Gesetz zum automatisierten Fahren), it is seen that a similar approach is taken by basically stipulating the responsibility of the ‘Auto Halter’ and supporting the responsibility structure of the ‘Auto Halter’ of the autonomous vehicle through a technical supervisor. At least in the transition period before strong artificial intelligence appears, this approach seems reasonable. It is necessary to systematically improve Korea's existing liability laws and laws related to artificial intelligence robots, and also continuous research on strong artificial intelligence is required.