Abstract

This report monitors evolving trends in the discourse and governance surrounding the adverse impacts of AI, encompassing
issues such as infringements on freedom and human dignity, discrimination, the propagation of misinformation and disinformation, and the potential existential threat posed by superintelligence. The Republic of Korea has actively engaged in global dialogues on AI governance, embodying its status as a digital democracy by incorporating outcomes into domestic regulations and policies. In December 2019, it unveiled the National Strategy for Artificial Intelligence, aligning with the OECD AI Principles from May 2019. In June 2020, it became a founding member of the GPAI, participating the AI Safety Summit in November 2023 as a host country for the next summit and sharing its policy initiative, Digital Bill of Rights.

The regulation of AI reveals a clash between digital democracy and digital authoritarianism. Even within digital democracies, divergence exists, with the United States prioritizing markets and Europe prioritizing individual rights. Anticipated is an acceleration in cooperation between digital democracies and authoritarians on AI regulation if an international organization dedicated to AI governance emerges. The potential roles for such an AI international organization range from being a standard-setting body akin to ICAO, a monitoring entity like
the IAEA, a collaborative research institution resembling CERN, to an assessment report organization like the IPCC.

The U.S. AI Executive Order marks a turning point from soft to hard regulation. Concurrently, Japan's announcement on the G7's consensus on "International Guidelines and Code of Conduct for Advanced Artificial Intelligence (AI) Development Organizations," based on the "Hiroshima AI Process," signifies a soft regulatory approach. The UK's subsequent declaration to convene an AI Safety Summit to deliberate on pre-validation of generative AI suggests a policy trajectory embracing a blend of hard and soft regulations.

The Republic of Korea must navigate a judicious balance between soft and hard regulation in its AI legislation. Incorporating a degree of hard regulation is imperative for establishing dependable AI systems and maximizing innovation. Additionally, careful consideration is warranted to assess whether certain AI development activities necessitate ex-ante international regulation.