UN의 AI 관리 방안 및 우리나라 정책에 대한 시사점 – UN 총회 결의안(A/78/L.49)과 UN AI 자문기구 최종 보고서를 중심으로 -

Artificial intelligence(AI) technology continues to advance at an unprecedented rate. Recognizing the significant risks that lie behind the positive potential and opportunities of this new technology, it is crucial to understand the ongoing discussions within the international community regarding response measures and for our nation to prepare accordingly. While AI can accelerate the achievement of the Sustainable Development Goals(SDGs) and enhance economic efficiency, it also carries the risks of human rights violations, a deepening digital divide, biased information delivery, and even concerns of human replacement or extinction. Against this backdrop, the international community, including organizations such as the Organisation for Economic Co-operation and Development(OECD), the G20, the United Nations Educational, Scientific and Cultural Organization(UNESCO), and the G7, has continued to discuss AI governance through policy proposals. In March 2024, the UN General Assembly set a significant milestone in AI governance discussions by adopting a US-led resolution on AI without a vote. The core of this resolution is the need to manage the risks of AI and contribute to achieving the SDGs by building safe, secure, and trustworthy AI systems. It also prioritizes bridging the gap between and within countries to ensure the benefits of AI's potential are shared equitably, emphasizing the need for global cooperation and the establishment of international governance. The resolution defines AI as a human-centric, reliable, and ethical technology in the non-military domain, clearly acknowledges both its potential and the risks of its misuse, and calls for technological and financial support for developing countries, as well as their expanded participation in governance. However, the resolution has also faced criticism for the ambiguity of its definition of AI, a lack of specificity in linking it to the SDGs, insufficient practical measures for supporting developing countries, and a failure to adequately warn of the dangers of superintelligence. Subsequently, in September 2024, the UN Secretary-General's High-level Advisory Body on AI published its final report, titled “Governing AI for Humanity.” Like the resolution, this report acknowledges the positive potential of AI while also warning of its risks, and it stresses the need to establish comprehensive and equitable global AI governance. In particular, it points out that current AI governance discussions are limited to certain countries and institutions and presents seven specific recommendations to bridge the gap and create more representative governance. The key recommendations are as follows: 1. The creation of an independent scientific panel to provide scientific knowledge and information for a common understanding of AI capabilities, opportunities, risks, and uncertainties. 2. The implementation of a policy dialogue on AI governance to promote international cooperation. 3. The establishment of a standards exchange to enhance international interoperability and information sharing. 4., 5., 6. The creation of a capacity-building network, a global AI fund, and a global AI data framework to close the AI gap for developing countries. 7. The establishment of an AI Office within the UN Secretariat to leverage the UN's capacity for the overall governance of AI. In conclusion, both documents recognize the dual nature of AI's potential and risks, emphasizing that bridging the AI capacity gap between developed and developing countries and ensuring the equitable sharing of benefits are key drivers for achieving the SDGs. Therefore, South Korea should actively engage by supporting capacity-building networks to close the AI gap, assessing and strengthening its domestic AI capabilities, taking a leading role in international AI standard-setting processes such as those at the ITU, and seeking an active role in international organizations like the GPAI.