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

作　　者：陆卫明 张文 王明 LU Weiming;ZHANG Wen;WANG Ming(School of Marxism,Xi’an Jiaotong University,Xi’an 710049,China;School of Marxism,Dongbei University of Finance and Economics,Dalian 116025,China)

机构地区：[1]西安交通大学马克思主义学院,陕西西安710049 [2]东北财经大学马克思主义学院,辽宁大连116025

出　　处：《西安交通大学学报(社会科学版)》2025年第2期63-77,共15页Journal of Xi'an Jiaotong University:Social Sciences

基　　金：教育部哲学社会科学研究重大课题攻关项目(24JZD040);全国高校思政课名师工作室项目(21SZJS61010698)。

摘　　要：以生成式人工智能为标志的新一代人工智能技术应用诱发的风险具有新生性、复杂性、跨境性等多重属性,加强全球人工智能治理、推动其向善且负责任发展已成为“共识”。这一“共识”的形成受多重动力驱使:各国政府前瞻布局以规避科技伦理潜在风险、人工智能企业亟须平衡风险治理与商业价值最优配置、国际组织聚焦地缘博弈与全球稳定、学界聚焦人工智能可持续性发展。“共识”的形成为构建全球人工智能“共治”机制提供了前提。全球人工智能治理从“共识”到“共治”面临着“囚徒困境”恶化引发的协作瓦解与零和博弈危机的根本性挑战、技术异步发展与治理标准不统一引发的双重性阻碍、文化价值冲突与伦理排序差异引发的适应性挑战、技术博弈与利益分配矛盾导致的协同性阻碍挑战。未来还需构建全球人工智能“共治”机制作为保障性路径,具体包括整合技术与价值的平衡机制、打造利益相关者协同治理生态、构建动态调整治理机制。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

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 

分 类 号：D815[政治法律—国际关系] TP18[政治法律—政治学]