AI Ethics: Bridging East-West Tech Governance, Examining through an Asian Lens

• In Event: Artificial Intelligence and Challenges in Policymaking

Abstract

This article critically explores the purported dominance of Western-centric perspectives in AI ethics and policy, questioning whether the evolving global AI governance regime is inadvertently sidelining Asian contexts. This study endeavors to evaluate AI ethics discourse internationally, but rooted in a critical examination of ethics in an Asian context, and looking beyond the more common focus on China or US-China relations. Our methodology involves a content analysis of AI ethics guidelines within major national policy frameworks, comparing those from the US and EU with those from Korea and Japan. This comparative approach, contrasting Western powers and (East) Asian countries, aims to identify political, social, and legal backgrounds characterizing individual Asian countries and their approach to AI ethics, with particular emphasis on Korea and Japan, contrasted against the US and EU. In order to provide insight on similarities, differences, and gaps, we perform a comparative study, drawing upon prominent national-level AI ethics guidelines to investigate how each country’s differing national values, and political and social backgrounds are presented in key national documents.

To gain a comprehensive understanding of AI ethics guidelines embodying diverse values, our research framework employs a qualitative comparative analysis. With a specific focus on Korea and Japan, we compare these with frameworks from key Western actors, namely the US and EU. Additionally, to delve into the nuances conveyed within each government's political and social landscape, we conduct an in-depth comparative analysis of specific components within the AI ethics guidelines. This scrutiny includes critical aspects such as Purpose, Goals or Visions, Principles, and Requirements, offering valuable insights into nuanced approaches adopted by governments when addressing AI ethics concerns.

Our findings suggest that the discourse at the global level, primarily shaped by Western nations and encapsulated in the globally-significant EU AI Act, inadequately represents Asian regions' needs. This disparity becomes evident in discussions of algorithmic bias centered on Western racial concepts, which do not align fully with Asian social, political, and legal frameworks.
More specifically, the comprehensive content analysis of AI ethics guidelines from the US, EU, Korea, and Japan reveals both common themes and distinct differences in how these nations approach the ethical considerations of AI. Despite variations in tone and methodological arguments, there is a shared overarching objective among these countries: to ensure that AI technologies are developed and deployed in ways that uphold human rights, safeguard individual privacy, promote fairness, and minimize harm. Another shared element is the recognition of the need for stakeholder engagement, emphasizing the importance of involving various actors in shaping the ethical framework for AI. This approach reflects a commitment to democratic and inclusive decision-making processes, allowing diverse perspectives to influence the development of AI technologies.

However, the nuances in their respective approaches are significant, providing valuable insights into the cultural, societal, and policy priorities of each nation. While the U.S. focuses on individual responsibility, the EU emphasizes environmental and societal considerations. Korea prioritizes community well-being, and Japan underscores social participation and the creation of a supportive environment. These differences highlight the rich tapestry of approaches to AI ethics worldwide, reflecting the diverse cultural, societal, and policy contexts in which AI technologies are deployed.

The analyses affirm literature highlighting a surge in a specific set of global AI ethics principles, including privacy, accountability, transparency, fairness, and safety. The discourse imbalance, emerging from Western governance, sidelines non-Western perspectives, such as the significance of human-centered values like equality, dignity, and community. In conclusion, the article advocates for a more inclusive, culturally attuned AI ethics framework.

Author

• Chee Hae Chung, Seoul National University