在人工智能（AI）技术迅猛发展的今天，全球正面临着前所未有的挑战与机遇。随着大型模型如ChatGPT和Sora的出现，AI正逐步接近所谓的“技术奇点”。然而，AI的巨大潜力也伴随着深伪、欺诈和自主武器系统等重大风险。在这一关键时刻，构建一个协调一致的全球AI治理框架变得尤为迫切。
清华大学人工智能国际治理研究院副院长梁正认为，AI的发展已经超越了国界，成为了一个需要全球共同应对的复杂问题。各国需要携手合作，以消除AI带来的风险，并确保其造福人类而非失控。他强调，中国在AI治理方面的独特视角和贡献，为全球提供了一个更加开放、公平和有效的治理路径。本文将深入探讨中国在AI治理方面的理念和实践，以及其对全球AI治理机制的影响。
-本文首发于《China Daily》-
原文链接:https://www.chinadaily.com.cn/a/202502/06/WS67a3ef2ea310a2ab06eaa572.html
核心内容：
1. AI技术的快速发展：
•AI技术进入了一个关键阶段，大模型如ChatGPT和Sora的突破性进展使AI接近所谓的“技术奇点”。
•AI的潜力巨大，但也伴随着深度伪造欺诈和自主武器系统等重大风险。
2. 全球治理的必要性：
•AI技术扩散超越国界，带来全球性的机遇和风险，需要各国合作来消除这些风险。
•人工智能能力呈指数级扩展、大模型自主性持续增强，现有治理架构已无法跟上技术发展的步伐，迫切需要建立更具适应性的治理框架。
3. 人工智能国际合作的努力：
•2023年，多个国家和地区开始重新调整其AI战略，英国和韩国分别举办了全球AI安全峰会。
•美国、加拿大、新加坡和日本等国家建立了AI安全研究机构，联合国也推动国际AI治理问题的进展。
•加强人工智能治理与合作的国际共识日益增强，为构建更统一、安全的全球体系奠定基础。
4. 当前治理机制的障碍：
•西方国家出于意识形态考虑，创建了限制更广泛国际参与的“排他圈子”，削弱了讨论的有效性和信任。
•发展中国家特别是全球南方国家面临基础设施不足、技术资源缺乏和人才短缺等问题，更容易受到AI带来的社会和经济冲击。
•全球AI治理体系碎片化，各国竞争标准制定与国际规则的主导权，阻碍了有效的国际合作。
5. 中国的AI治理理念：
•中国作为AI技术和应用的全球领导者，以负责任的态度应对AI治理挑战，提出包容和有效的治理理念。
•在国内，中国通过法律框架、技术标准和伦理指南的三元治理促进AI健康发展。
•在国际上，中国坚持人类命运共同体的理念立场，推动多边合作治理，提出《全球人工智能治理倡议》，支持在联合国框架内建立开放和包容的人工智能治理机制。
6. 平衡发展与安全：
•中国强调发展与安全的平衡，认为“不发展就是最大的不安全”。
•支持联合国关于人工智能能力建设国际合作的决议，提出《人工智能能力建设普惠计划》，帮助缩小全球AI发展差距，促进包容和平衡的发展。
7. 敏捷治理模式：
•中国探索敏捷治理模式，强调跨部门合作，政府、科学界和行业共同管理风险，提高安全性。
•鼓励分享最佳实践，促进合作解决方案，创建响应迅速、适应AI快速发展的治理系统。
China shaping AI governance mechanism
The development of artificial intelligence has entered a pivotal phase. With groundbreaking advancements in large models such as ChatGPT and Sora, AI is approaching what has been termed as "technological singularity".The allure of AI's potential is undeniable, but its immense potential is accompanied by significant risks including deepfakes, frauds and autonomous weapons systems.
The complexities and interconnectedness of AI pose a new global challenge. Hence, building a coordinated global governance framework for AI is no longer optional; it is an urgent necessity.
AI transcends national boundaries, creating both global opportunities and risks that no country alone can manage. Hence, countries across the world need to work together to eliminate the risks.
The urgency to create an adaptive governance framework is underscored by the rapid, unpredictable growth of AI. Given that AI capabilities are expanding exponentially and large models are becoming increasingly autonomous, the existing AI governance structures can no longer keep pace with the developments. And in the absence of proactive frameworks to deal with emerging risks, AI could outpace regulations, leading to catastrophic consequences.
Therefore, we need to act now, in order to ensure AI benefits humankind rather than spiraling out of human control and harming society.
This is a moment of reckoning for the global community. In 2023, countries and regions across the world began recalibrating their AI strategies, with the United Kingdom hosting the first global summit on AI safety in November 2023 and the Republic of Korea hosting the second in May 2024.
Besides, countries such as the United States, Canada, the UK, Singapore and Japan have set up AI safety research institutes to foster international coordination on critical issues. At the heart of these efforts is the United Nations, which has pushed forward the issue of international governance of AI. In fact, the UN established a high-level AI advisory body in October 2023.
These developments reflect the growing international consensus on strengthening AI governance and cooperation, laying the groundwork for a more integrated, secure global system. But as countries and regions endeavor to establish a comprehensive AI governance mechanism, several obstacles associated with the current mechanism — including the lack of openness, fairness and effectiveness of the current governance processes — stand out.
First, Western countries, driven by their ideological considerations, have created "exclusive circles" that limit broader international participation. This has turned the discussions on AI governance into closed, exclusive affairs which undermines the effectiveness of the discussions and hinders the ability of many countries to engage in the discussions, eroding the trust necessary to reach a global consensus.
Second, many countries in the Global South face serious challenges, including limited infrastructure, lack of technical resources and shortage of talent, making them more vulnerable to the social and economic disruptions caused by AI. As a result, such countries face greater risk of being left behind in the global race of AI.
Third, the existing global AI governance system is fragmented, with countries competing with each other to influence global AI standards and governance. This fragmentation hampers meaningful global cooperation, and weakens the global AI governance framework.
In this context, China's approach to AI governance offers a unique perspective. As one of the global leaders in both AI technology and application, China approaches the challenges of AI governance with a sense of responsibility, and follows a philosophy that can guide the world toward a more inclusive and effective governance system.
Domestically, this philosophy is reflected in a robust legal framework, high technical standards and ethical guidelines for the healthy development of AI. Internationally, to promote multilateral cooperative governance, China has proposed programs such as the Global AI Governance Initiative. China's contribution to the discussions on global AI governance is not simply theoretical; it reflects "Eastern wisdom" which can facilitate open, fair and effective AI governance.
Committed to the idea of a shared destiny for mankind, China views AI governance as a field of cooperation rather than competition. By advocating for the establishment of an open global dialogue platform and inclusive governance mechanism within the UN framework, China supports multilateralism, and promotes pragmatic cooperation and fair participation among countries.
Also, China emphasizes the importance of balancing development with security. Recognizing that AI governance is not only about addressing risks but also about meeting the developmental needs of countries, China stresses that the failure to achieve development constitutes the greatest insecurity. That's why it supports UN initiatives for AI capacity building, and has proposed the "AI Capacity-Building Action Plan for Good and for All "to help bridge the global AI divide and foster inclusive and balanced global development.
China is exploring a flexible governance model to address AI's unique challenges by laying emphasis on collaboration across sectors, bringing together governments, the scientific community and industries to collectively manage risks and improve safety. By encouraging the sharing of best practices and fostering cooperative solutions, China aims to create a governance system that is responsive and adaptable to the fast-evolving AI landscape.
Through these contributions, China is laying the groundwork for a more open, fair and effective international AI governance mechanism. At a time when the world is grappling with the implications of AI, China offers a path that prioritizes collective security, balanced development and a shared vision for the future of this transformative technology.
（作者：梁正，清华大学人工智能国际治理研究院副院长、人工智能治理研究中心主任、中国科技政策研究中心副主任、公共管理学院教授）
来源 | China Daily
