Discussion on the Ethical Relationship between Humans and Artificial Intelligence: A Multi-stakeholder Governance Perspective

Rixin College, Tsinghua University, Beijing, China

^* Corresponding author: yuanyr22@mails.tsinghua.edu.cn

Abstract

The rapid advancement of large language models (LLMs) and other artificial intelligence (AI) technologies has profoundly impacted our society. While existing governance frameworks address network and data security, the ethical governance of AI remains underdeveloped and cannot be solely reliant on past experiences. This article explores the human-AI relationship through a people-first ethical lens, focusing on the responsibilities of three key stakeholders: developers, regulators, and end- users. We first delineate the fundamental differences between AI and traditional automation systems. Then, we systematically review four critical ethical issues in AI governance—privacy and data security, algorithmic bias, responsibility attribution, and emotional and ethical concerns—grounding each in existing literature. Subsequently, we propose a multi-stakeholder interaction framework, incorporating technical and policy case studies to illustrate practical challenges and solutions. We conclude that building a secure, reliable, and controllable AI ecosystem requires the concerted efforts of all parties, with continuous public discourse defining the boundaries of our values.