South Korean Public Value Coproduction Towardsg AI for Humanity': A Synergy of Sociocultural Norms and Multistakeholder Deliberation in Bridging the Design and Implementation of National AI Ethics Guidelines

Abstract

As emerging technologies such as Big Data, Artificial Intelligence (AI), robotics, and the Internet of Things (IoT) pose fundamental challenges for global and domestic technological governance, theg Fourth Industrial Revolution' (4IR) comes to the fore with AI as a frontrunner, generating discussions on the ethical elements of AI amongst key stakeholder groups, such as government, academia, industry, and civil society. However, in recent AI ethics and governance scholarship, AI ethics design appears to be divorced from AI ethics implementation, an implicit partition that results in two separate matters of theory and practice, respectively, and thus invokes efforts to bridge theg gap' between the two. Such a partition potentially overcomplicates the discussion surrounding AI ethics and limits its productivity. This paper thus presents South Korea's people-centeredg National Guidelines for Artificial Intelligence Ethics'(êμ ê°€ì ê3μì§€ë ¥ì Currency signë▪¬ê °ìCurrency sign€;g Guidelines') and their development under the Moon administration as a case study that can help readers conceptualize AI ethics design and implementation as a continuous process rather than a partitioned one. From a public value perspective, the case study examines the Guidelines and the multistakeholder policymaking infrastructure that serves as the foundation for both the Guidelines' design and implementation. This examination draws from literature in AI ethics and governance, public management and administration, and Korean policy and cultural studies as well as government and public documents alongside 9 interviews with members from the four stakeholder groups that collectively designed and continue to deliberate upon the Guidelines. Further, the study specifically focuses on (i) identifying public values that were highlighted by the Guidelines, (ii) investigating how such values reflect prevalent Korean sociocultural norms, and (iii) exploring how these values, in a way made possible by Korean sociocultural norms and policymaking, have been negotiated amongst the four stakeholder groups in a democratic public sphere to be ultimately incorporated into the Guidelines and prepared for implementation. This paper hopes to contribute to theory-building in AI ethics and provide a point of comparison in the international stage for future research concerning AI ethics design and implementation.