국제적 권고ㆍ지침 분석에 바탕을 둔 “인공지능(AI) 윤리” 관련 국제조약안의 모색

The purpose of the article is to find out the framework of an international treaty on the ethics of Artificial Intelligence (AI). To this end, first, this article analyzes the main content of existing relevant international documents, namely “Ethics Guidelines for Trustworthy AI” adopted by the European Commission in 2019, “Recommendation of the Council on Artificial Intelligence” adopted by the OECD in 2019, “Recommendation on the Ethics of AI" by the UNESCO in 2021 and documents adopted by civic groups. Secondly, upon the above works, it extracts common and main principles that should be respected through the entire life cycle of AI: namely, the protection of human rights, democracy, the rule of law and distributive justice. Thirdly, we can identify the specific provisions that should be explicitly mentioned in the future treaty. These are as follows: provisions that reaffirm human dignity and fundamental human rights, provisions concerning the implementation of the obligations of the treaty into domestic legislations and rules, provisions to guarantee the transparency of AI ​​system, provisions to establish the ethical impact assessment. A provision regarding appropriate remedies for victims can be also deemed necessary. Concerning the reaffirmation of human dignity and fundamental human rights, the general scheme of an international treaty on the ethics of AI should be based on the fundamental principles stipulated in the 1948 Universal Declaration of Human Rights and international conventions on the protection and promotion of human rights to which almost States have ratified. However, the treaty should be modified somewhat considering the specificity of the AI system. In other words, the duty to protect and respect human rights throughout the life cycle of the AI ​​system and to fulfill treaty obligations might include not only State but also individuals who participate in the development and operation of the AI ​​system. Again, two contradictory aspects of AI system should be emphasized: first, it can contribute to the prosperity and well-being of human society through development and innovation, but on the other hand, the malfunctioning of AI system can pose a potential threat to the protection of human rights, respect for the spirit of democracy and the rule of law. Accordingly, if the international community can adopt in the near future a international treaty on the ethics of AI ​​system, the related legal issues can be dealt with in a more transparent and appropriate way.