Author
Ms Xinyu Xie
Organisation/Institution
The University of Hong Kong, Faculty of Law
Country
HONG KONG (SAR OF CHINA)
Panel
Law and Society
Title
Navigating AI Value Conflicts: A "Harmony in Diversity" Approach
Abstract
The foundational goal of AI value alignment is to ensure that intelligent agents operate in harmony with human values. This paper moves beyond prevailing technical debates to address a more fundamental normative dilemma: Which values should AI be aligned with, and how should conflicts among them be adjudicated in specific contexts? Theoretical and empirical evidence consistently reveals deep divergences in moral principles, cognitive frameworks, and societal expectations, rendering the pursuit of a universal alignment standard not only difficult but arguably untenable. Confronted with this challenge, Asia’s profound cultural diversity and historical experience in managing complex differences offer valuable inspiration. It argues that philosophical principles such as “harmony in diversity” (和而不同) and “seeking common ground while reserving differences” (求同存异) can inform the construction of a deliberative order for value negotiation. Grounded in deliberative constructivism, procedural justice, and discourse theory, the study proposes a potential paradigm shift from the pursuit of substantive alignment to the implementation of procedural governance. Rather than pursuing an elusive, universally “correct” set of values, this approach acknowledges value pluralism as its foundation, abandons the quest for perfect substantive alignment, and focuses instead on establishing differentiated and contextualized governance rules from a procedural perspective. Ultimately, this approach is grounded in the rule of law, which provides a credible framework—through baseline standards and procedural safeguards—for the coexistence, dialogue, and negotiation of diverse values.
Biography
Xie Xinyu is currently a second-year PhD student at the Faculty of Law, the University of Hong Kong (HKU). Her doctoral research focuses on administrative law and empirical legal studies, with a particular interest in the intersection of law, governance, and technology. She has experience in both doctrinal and empirical methodologies, including statistical analysis and comparative approaches, and has presented her work at international conferences on law and artificial intelligence(e.g., China University of Political Science and Law's “Discretionary Justice in the Digital Age” and Tsinghua University's International Forum on Computational Law). Her recent projects explore questions of procedural justice, responsible AI governance, and the role of empirical evidence in shaping administrative decision-making. Through her interdisciplinary background, she aims to contribute to a deeper understanding of how legal frameworks can adapt to emerging challenges in governance and technology while maintaining fairness and accountability.
