西安交通大学学报（社会科学版）2025，Vol.45Issue(2)：63-77,15.DOI:10.15896/j.xjtuskxb.202502006
全球人工智能治理从"共识"到"共治":动力、挑战与路径
From"Consensus"to"Co-Governance"in Global AI Governance:Drivers,Challenges,and Pathways
摘要
Abstract
The exploration of an effective path for global AI governance has emerged as a pivotal subject in academic discourse.The risks associated with the implementation of new-generation AI technology,exemplified by generative AI,are characterized by novel,intricate and transnational characteristics,necessitating a collaborative approach that transcends individual nation or AI enterprise capabilities.This imperative underscores the need for the augmentation of global cooperation and the establishment of a mechanism for shared governance.The establishment of such a mechanism is a fundamental guarantee for ensuring the inclusive and sustainable development of artificial intelligence technology.Evidence of an emerging consensus on strengthening global AI governance is apparent from the Global AI Governance Initiative proposed by President Xi Jinping at the Third Belt and Road Forum for International Cooperation in 2023 to the Statement on the Development of Inclusive and Sustainable AI for the Benefit of Humanity and the Planet at the AI Action Summit in Paris,France,in early 2025.By the time of the AI Action Summit in Paris,France in early 2025,the Statement on the Development of Inclusive and Sustainable Artificial Intelligence for the Benefit of Humanity and the Planet had proven that a preliminary consensus had been formed on strengthening global AI governance.In order to promote AI technology that does not harm humans and creates more benefits for them,it is necessary to promote global AI governance from consensus to shared governance. The present study draws on a series of bills and reports,including the EU's Artificial Intelligence Act and China's Global AI Governance Report,in order to examine the risks induced by the application of AI technology and the current status of global governance from the consensus that AI technology always creates more well-being for human beings.The study finds that strengthening global AI governance has become a consensus among governments,AI enterprises,academics,and the public.The study further elucidates that the efficacy with which global AI risks can be managed is contingent upon the establishment of a global AI shared governance mechanism.This transition necessitates a comprehensive understanding of the challenges associated with shifting from a consensus to a shared governance'model,while leveraging the foundations laid by the initial consensus.To facilitate this transition,it is imperative to take a pivotal step from consensus to shared governance,employing innovative governance approaches to establish a global framework for the collective management of AI. The study has reached the following four conclusions:(1)the formation of a consensus on global AI governance is driven by multiple forces;(2)The Prisoner s Dilemma is a fundamental challenge for global AI governance to transition from consensus to common governance;(3)Achieving the pivotal step of global AI governance from consensus to common governance is a key challenge for the establishment of a common governance for global AI;(4)The fourth point emphasizes that the establishment of a global AI governance mechanism is paramount to prevent the unregulated proliferation of AI technology. The governance of AI on a global scale necessitates not only the formulation of pertinent policies,laws and regulations,but also the contributions of the academic community.The study proffers the academic perspective that global AI must be realized from consensus to common governance until a global AI common governance mechanism is established,which is of great significance to the promotion of the good development of AI and the protection of human welfare.This assertion underscores the pivotal role that academic expertise and discourse play in shaping the governance of global AI,underscoring the necessity for dialogue and collaboration with relevant experts and scholars to ensure the integration of diverse perspectives in the formulation of policies and the establishment of a robust governance framework.关键词
全球人工智能治理/负责任人工智能/人工智能技术/囚徒困境/合作博弈/"共治"机制Key words
global AI governance/responsible AI/AI technology/prisoner's dilemma/cooperative game/co-governance mecha-nism分类
政治法律引用本文复制引用
陆卫明,张文,王明..全球人工智能治理从"共识"到"共治":动力、挑战与路径[J].西安交通大学学报（社会科学版）,2025,45(2):63-77,15.基金项目
教育部哲学社会科学研究重大课题攻关项目(24JZD040) （24JZD040）
全国高校思政课名师工作室项目(21SZJS61010698). （21SZJS61010698）
