Tuesday, August 06, 2024

为全球人工智能治理提供中国方案 [Providing a Chinese solution for global AI governance] (1 August 2024) 中国社会科学网 (China Social Science Network) Policy Papers

 


 

 当前,全球人工智能技术快速发展,对经济社会发展和人类文明进步产生深远影响,给世界带来巨大机遇。与此同时,人工智能技术也带来难以预知的各种风险和复杂挑战。人工智能治理攸关全人类命运,是世界各国面临的共同课题。2023年10月18日,习近平主席在第三届“一带一路”国际合作高峰论坛开幕式主旨演讲中提出了“全球人工智能治理倡议“,系统回答了为什么要加强、如何加强人工智能全球治理的重大时代课题,提出了中国方案,贡献了中国智慧,为凝聚国际社会共识、推进全球共享共治发挥了积极作用。[At present, the rapid development of global artificial intelligence technology has a profound impact on economic and social development and the progress of human civilization, bringing huge opportunities to the world. At the same time, artificial intelligence technology also brings various unpredictable risks and complex challenges. Artificial intelligence governance is related to the fate of all mankind and is a common issue faced by countries around the world. On October 18, 2023, President Xi Jinping proposed the "Global Artificial Intelligence Governance Initiative" in his keynote speech at the opening ceremony of the third "Belt and Road" International Cooperation Summit Forum. He systematically answered the major contemporary issue of why and how to strengthen the global governance of artificial intelligence, proposed a Chinese solution, contributed Chinese wisdom, and played a positive role in building consensus in the international community and promoting global sharing and co-governance.] (中国社会科学网 (China Social Science Network): 为全球人工智能治理提供中国方案 [Providing a Chinese solution for global AI governance])

Every major power will have its AI law, and more generally its core premises around which AI legalities may be woven. The European Union has given itself both (EU AI Act here; Ethics guidelines for trustworthy AI here). The United States has developed its premises and remains committed to markets based guided legalities (Blueprint for an AI Bill of Rights here; also here, here and here). China continues to develop its own normative framework and legalities, now in the form of a "Global Artificial Intelligence Governance Initiative" [“全球人工智能治理倡议“]. These are meant not merely to discipline the use and development of AI within a State but also to produce the normative narrative framework against which all global efforts will be undertaken. As such, each reflects the political-economic models of their origins and the structures for their application elsewhere.

As such, all three seek to project their frameworks, normative narratives, and their underlying structures of control outward. In this sense one can understand AI law and norms as effectively the framing of platforms through which  AI related operations can be coordinated and undertaken, while at the same time serving as a border/barrier that protects the integrity of one AI normative platform from others.  A system of multi-platform platform autopoiesis emerges. The object remains the same across platforms:我们重申,各国应在人工智能治理中加强信息交流和技术合作,共同做好风险防范,形成具有广泛共识的人工智能治理框架和标准规范,不断提升人工智能技术的安全性、可靠性、可控性、公平性。[We reiterate that all countries should strengthen information exchange and technical cooperation in AI governance, jointly prevent risks, form an AI governance framework and standard specifications with broad consensus, and continuously improve the security, reliability, controllability and fairness of AI technology. ]. But like the others, these systems are meant to produce the illusion of control and instrumentalization of AI when in fact they reflect  an exteriorized manifestation of efforts at self control even as the human becomes the instrument of AI based systemicities (for my extended analysis see 'The Soulful Machine, the Virtual Person, and the “Human” Condition').

To these ends China has devoted resources to the development of an A regulatory narrative with Chinese characteristics.  These are reflected in five highlighted essays:

1. 防范风险是全球人工智能治理的主要日标; 郎平; 中国社会科学院世界经济与政治研究所研究员 (Preventing risk is the main goal of global AI governance, Lang Ping, Researcher at the Institute of World Economics and Politics, Chinese Academy of Social Sciences)

2. 人工智能治理在于平衡安全与发展; 李艳; 中国现代国际关系研究院科技与网络安全研究所执行所长、研究员 (AI governance is about balancing security and development; Li Yan, Executive Director and Researcher of the Institute of Science, Technology and Cybersecurity, China Institutes of Contemporary International Relations)

3. 《全球人工智能治理倡议》中的中国智慧鲁传颖; 上海国际问题研究院研究员、网络空间国际治理研究中心秘书长 (Chinese Wisdom in the Global AI Governance Initiative; Lu Chuanying; Researcher at the Shanghai Institutes for International Studies, Secretary-General of the Center for International Cyberspace Governance Research)

4. 《全球人工智能治理倡议》为数子公平贡献中国方案; 全燕; 广东外语外贸大学新闻与传播学院教授 ("Global AI Governance Initiative" contributes Chinese solutions to digital equity; Quan Yan; Professor of the School of Journalism and Communication, Guangdong University of Foreign Studies)

5. 全球人工智能治理倡议的世界意义; 费艳颖, 大连理工大学公共管理学院教授; 赵凯, 大连理工大学马克思主义学院博士研究生 (The global significance of the global artificial intelligence governance initiative; Fei Yanying, Professor of the School of Public Administration, Dalian University of Technology; Zhao Kai, PhD candidate of the School of Marxism, Dalian University of Technology)

Each provides useful insights into the development of the "Global Artificial Intelligence Governance Initiative" [“全球人工智能治理倡议“], which follows below in the original Chinese and in a crude English translation along with the text of the five essays identified above.

全球人工智能治理倡议

2023-10-20 15:12

人工智能是人类发展新领域。当前,全球人工智能技术快速发展,对经济社会发展和人类文明进步产生深远影响,给世界带来巨大机遇。与此同时,人工智能技术也带来难以预知的各种风险和复杂挑战。人工智能治理攸关全人类命运,是世界各国面临的共同课题。

在世界和平与发展面临多元挑战的背景下,各国应秉持共同、综合、合作、可持续的安全观,坚持发展和安全并重的原则,通过对话与合作凝聚共识,构建开放、公正、有效的治理机制,促进人工智能技术造福于人类,推动构建人类命运共同体。

我们重申,各国应在人工智能治理中加强信息交流和技术合作,共同做好风险防范,形成具有广泛共识的人工智能治理框架和标准规范,不断提升人工智能技术的安全性、可靠性、可控性、公平性。我们欢迎各国政府、国际组织、企业、科研院校、民间机构和公民个人等各主体秉持共商共建共享的理念,协力共同促进人工智能治理。

为此,我们倡议:

——发展人工智能应坚持“以人为本”理念,以增进人类共同福祉为目标,以保障社会安全、尊重人类权益为前提,确保人工智能始终朝着有利于人类文明进步的方向发展。积极支持以人工智能助力可持续发展,应对气候变化、生物多样性保护等全球性挑战。

——面向他国提供人工智能产品和服务时,应尊重他国主权,严格遵守他国法律,接受他国法律管辖。反对利用人工智能技术优势操纵舆论、传播虚假信息,干涉他国内政、社会制度及社会秩序,危害他国主权。

——发展人工智能应坚持“智能向善”的宗旨,遵守适用的国际法,符合和平、发展、公平、正义、民主、自由的全人类共同价值,共同防范和打击恐怖主义、极端势力和跨国有组织犯罪集团对人工智能技术的恶用滥用。各国尤其是大国对在军事领域研发和使用人工智能技术应该采取慎重负责的态度。

——发展人工智能应坚持相互尊重、平等互利的原则,各国无论大小、强弱,无论社会制度如何,都有平等发展和利用人工智能的权利。鼓励全球共同推动人工智能健康发展,共享人工智能知识成果,开源人工智能技术。反对以意识形态划线或构建排他性集团,恶意阻挠他国人工智能发展。反对利用技术垄断和单边强制措施制造发展壁垒,恶意阻断全球人工智能供应链。

——推动建立风险等级测试评估体系,实施敏捷治理,分类分级管理,快速有效响应。研发主体不断提高人工智能可解释性和可预测性,提升数据真实性和准确性,确保人工智能始终处于人类控制之下,打造可审核、可监督、可追溯、可信赖的人工智能技术。

——逐步建立健全法律和规章制度,保障人工智能研发和应用中的个人隐私与数据安全,反对窃取、篡改、泄露和其他非法收集利用个人信息的行为。

——坚持公平性和非歧视性原则,避免在数据获取、算法设计、技术开发、产品研发与应用过程中,产生针对不同或特定民族、信仰、国别、性别等偏见和歧视。

——坚持伦理先行,建立并完善人工智能伦理准则、规范及问责机制,形成人工智能伦理指南,建立科技伦理审查和监管制度,明确人工智能相关主体的责任和权力边界,充分尊重并保障各群体合法权益,及时回应国内和国际相关伦理关切。

——坚持广泛参与、协商一致、循序渐进的原则,密切跟踪技术发展形势,开展风险评估和政策沟通,分享最佳操作实践。在此基础上,通过对话与合作,在充分尊重各国政策和实践差异性基础上,推动多利益攸关方积极参与,在国际人工智能治理领域形成广泛共识。

——积极发展用于人工智能治理的相关技术开发与应用,支持以人工智能技术防范人工智能风险,提高人工智能治理的技术能力。

——增强发展中国家在人工智能全球治理中的代表性和发言权,确保各国人工智能发展与治理的权利平等、机会平等、规则平等,开展面向发展中国家的国际合作与援助,不断弥合智能鸿沟和治理能力差距。积极支持在联合国框架下讨论成立国际人工智能治理机构,协调国际人工智能发展、安全与治理重大问题。

Global AI Governance Initiative

2023-10-20 15:12

AI is a new field for human development. At present, the rapid development of global AI technology has a profound impact on economic and social development and the progress of human civilization, bringing huge opportunities to the world. At the same time, AI technology also brings various unpredictable risks and complex challenges. AI governance is related to the fate of all mankind and is a common issue faced by all countries in the world.

In the context of multiple challenges facing world peace and development, all countries should uphold a common, comprehensive, cooperative and sustainable security concept, adhere to the principle of giving equal importance to development and security, build consensus through dialogue and cooperation, build an open, fair and effective governance mechanism, promote AI technology to benefit mankind, and promote the building of a community with a shared future for mankind.

We reiterate that all countries should strengthen information exchange and technical cooperation in AI governance, jointly prevent risks, form an AI governance framework and standard specifications with broad consensus, and continuously improve the security, reliability, controllability and fairness of AI technology. We welcome governments, international organizations, enterprises, scientific research institutions, non-governmental organizations and individual citizens to uphold the concept of consultation, construction and sharing, and work together to promote AI governance.

To this end, we propose:

—— The development of artificial intelligence should adhere to the concept of "people-oriented", aim to enhance the common welfare of mankind, and ensure social security and respect for human rights and interests as the premise, to ensure that artificial intelligence always develops in a direction that is conducive to the progress of human civilization. Actively support the use of artificial intelligence to promote sustainable development and respond to global challenges such as climate change and biodiversity conservation.

—— When providing artificial intelligence products and services to other countries, we should respect the sovereignty of other countries, strictly abide by the laws of other countries, and accept the jurisdiction of other countries' laws. Oppose the use of artificial intelligence technology to manipulate public opinion, spread false information, interfere in other countries' internal affairs, social systems and social order, and endanger the sovereignty of other countries.

—— The development of artificial intelligence should adhere to the purpose of "intelligence for good", abide by applicable international law, conform to the common values ​​of peace, development, fairness, justice, democracy and freedom for all mankind, and jointly prevent and combat the abuse of artificial intelligence technology by terrorism, extremist forces and transnational organized crime groups. Countries, especially major powers, should take a cautious and responsible attitude towards the research and development and use of artificial intelligence technology in the military field.

—— The development of artificial intelligence should adhere to the principles of mutual respect, equality and mutual benefit. All countries, regardless of size, strength, and social system, have the right to develop and use artificial intelligence equally. Encourage the world to jointly promote the healthy development of artificial intelligence, share artificial intelligence knowledge, and open source artificial intelligence technology. Oppose drawing lines based on ideology or building exclusive groups to maliciously obstruct the development of artificial intelligence in other countries. Oppose the use of technology monopoly and unilateral coercive measures to create development barriers and maliciously block the global artificial intelligence supply chain.

—— Promote the establishment of a risk level test and assessment system, implement agile governance, classify and grade management, and respond quickly and effectively. R&D entities continue to improve the explainability and predictability of artificial intelligence, improve the authenticity and accuracy of data, ensure that artificial intelligence is always under human control, and create auditable, supervised, traceable, and trustworthy artificial intelligence technology.

—— Gradually establish and improve laws and regulations to protect personal privacy and data security in artificial intelligence research and development and applications, and oppose theft, tampering, disclosure and other illegal collection and use of personal information.

—— Adhere to the principles of fairness and non-discrimination, and avoid bias and discrimination against different or specific nationalities, beliefs, countries, genders, etc. in the process of data acquisition, algorithm design, technology development, product development and application.

——Adhere to ethics first, establish and improve AI ethical standards, norms and accountability mechanisms, form AI ethical guidelines, establish a science and technology ethics review and supervision system, clarify the responsibilities and power boundaries of AI-related entities, fully respect and protect the legitimate rights and interests of various groups, and promptly respond to relevant domestic and international ethical concerns.

——Adhere to the principles of broad participation, consensus, and gradual progress, closely follow the development of technology, conduct risk assessment and policy communication, and share best practices. On this basis, through dialogue and cooperation, on the basis of fully respecting the differences in policies and practices of various countries, promote the active participation of multiple stakeholders and form a broad consensus in the field of international AI governance.

——Actively develop relevant technology development and application for AI governance, support the use of AI technology to prevent AI risks, and improve the technical capabilities of AI governance.

——Enhance the representation and voice of developing countries in the global governance of AI, ensure equal rights, equal opportunities, and equal rules for the development and governance of AI in various countries, carry out international cooperation and assistance for developing countries, and continuously bridge the intelligence gap and governance capacity gap. Actively support the discussion of the establishment of an international AI governance institution under the framework of the United Nations to coordinate major issues of international AI development, security and governance.



防范风险是全球人工智能治理的主要目标
2024-05-17 作者:郎平 来源:中国社会科学网

理解《全球人工智能治理倡议》(以下简称“倡议”)需要从国际安全的角度去理解,比如说人工智能如何影响世界和平。在我看来,倡议从全球的视角来说,就是从全球治理的角度构建人类命运共同体。全球人工智能治理倡议考虑了两个维度,一个是人类的整体的生存和发展,还有一个就是国际安全。

  倡议属于全球人工智能治理的范畴,主要目的是为了规避风险。一个是伦理道德风险,另一个就是安全风险。所以比如说我们有的时候,我们每个国家怎么制定自己的人工智能的治理政策,这个是每个国家的主权管辖范围之内的事情。

  在大多数情况下,人工智能的政策均由国家制定。一国制定本国战略或者进行国际合作的时候,常常需要兼顾他国的反应,但这本质上是本国主权的事情,所以,我们在讨论全球治理的时候,特别是人工智能全球治理的时候,需要结合各国的政策,达成一个统一的标准,才能够完善全球人工智能治理的效果。

  倡议主要目的是防范人工智能带来的风险。风险既蕴含挑战也有机遇,当我们讨论风险的时候,人工智能它既带来了安全,就是发展的种种的这种赋能的效应,同时它也会带来很多我们说的安全问题,就比如说假新闻的问题、对隐私保护的问题、知识产权的问题。

  我们怎么去认识人工智能给我们带来的这些风险?首先,要把人工智能可能会带来的各种安全问题一一列举,然后去认识,比如:可以按照动机划分,风险的产生是无意的还是有意的,还有人为行动等。如果不是人为的,通过技术手段预防防范就可以了,但若是人为的,就不能采用这种技术手段去解决。所以就是说要区分他的动机,

  除了无意安全风险和有意安全风险外,对我们很重要的还有一个是主观安全威胁还是客观安全威胁。也就是说,在我们认识到它可能会给我们带来风险的时候,其实我们的认知里面是既包含了客观的部分,就是客观上人工智能在发展,也包含了主观上的认识。但主观上的认识可能是由于安全能力的不足,同样的客观的安全风险,对于安全能力强的国家来说,主观上就不觉得是个挑战。

  对于人工智能的风险而言,需要去区分它到底是一个客观的还是主观的,因为客观的是相对来说不变的,但是主观我们跟我们自身的很多因素相关,所以这也是我们认识这个风险应该怎么去做的一个前提,还有一个就是安全风险,它可能会给我们带来什么样的威胁,威胁的严重程度是威胁到了全局的利益,重大的利益,还是一般的利益,都需要去通过不同的维度去认识这些安全风险,然后才能去防范,我们需要用不同的应对方式来对应来防范风险。

  对于人工智能而言,我们需要提升安全能力,只要有能力,就可以拥抱风险,而不是惧怕风险,因为风险是无处不在的,所以我们要从风险里面去找到挑战,对于人工智能来说,从理论上来讲,要坚持总体国家安全观,统筹发展和安全,在安全的基础上,才能够实现治理。

  中国社会科学网李想/采访整理

  (作者系中国社会科学院世界经济与政治研究所国家安全研究室主任、研究员)

Preventing risks is the main goal of global AI governance
2024-05-17 Author: Lang Ping Source: China Social Sciences Network

To understand the "Global AI Governance Initiative" (hereinafter referred to as the "Initiative"), we need to understand it from the perspective of international security, such as how AI affects world peace. In my opinion, the Initiative, from a global perspective, is to build a community with a shared future for mankind from the perspective of global governance. The Global AI Governance Initiative takes into account two dimensions, one is the overall survival and development of mankind, and the other is international security.

The Initiative belongs to the category of global AI governance, and its main purpose is to avoid risks. One is ethical and moral risks, and the other is security risks. So for example, sometimes, how each of our countries formulates its own AI governance policies is within the sovereign jurisdiction of each country.

In most cases, AI policies are formulated by the country. When a country formulates its own strategy or conducts international cooperation, it often needs to take into account the reactions of other countries, but this is essentially a matter of national sovereignty. Therefore, when we discuss global governance, especially global AI governance, we need to combine the policies of various countries and reach a unified standard to improve the effectiveness of global AI governance.

The main purpose of the Initiative is to prevent the risks brought by AI. Risks contain both challenges and opportunities. When we discuss risks, artificial intelligence not only brings security, that is, the various enabling effects of development, but also brings many security issues, such as fake news, privacy protection, and intellectual property rights.

How do we understand the risks that artificial intelligence brings to us? First of all, we need to list the various security issues that artificial intelligence may bring, and then understand them. For example, we can classify them according to motivations, whether the risks are unintentional or intentional, and whether there are human actions. If it is not man-made, it can be prevented and prevented by technical means, but if it is man-made, this technical means cannot be used to solve it. So it is necessary to distinguish his motivations.

In addition to unintentional and intentional security risks, it is also very important for us to determine whether it is a subjective security threat or an objective security threat. In other words, when we realize that it may bring us risks, our cognition actually includes both the objective part, that is, the objective development of artificial intelligence, and the subjective cognition. However, the subjective cognition may be due to the lack of security capabilities. The same objective security risk is not considered a challenge for countries with strong security capabilities.

As for the risks of artificial intelligence, we need to distinguish whether it is objective or subjective, because the objective is relatively unchanged, but the subjective is related to many factors of our own, so this is also a prerequisite for us to understand how to deal with this risk. Another is security risk. What kind of threat may it bring to us? The severity of the threat is whether it threatens the overall interests, major interests, or general interests. We need to understand these security risks from different dimensions before we can prevent them. We need to use different response methods to prevent risks.

For artificial intelligence, we need to improve our security capabilities. As long as we have the ability, we can embrace risks instead of fearing them, because risks are everywhere, so we must find challenges from risks. For artificial intelligence, theoretically speaking, we must adhere to the overall national security concept, coordinate development and security, and achieve governance on the basis of security.

China Social Sciences Network Li Xiang/Interview and compilation

(The author is the director and researcher of the National Security Research Office of the Institute of World Economics and Politics of the Chinese Academy of Social Sciences)




李艳:人工智能治理在于平衡安全与发展
2024-05-11 作者:李艳 来源:中国社会科学网

2023年10月,习近平主席在第三届“一带一路”国际合作高峰论坛上提出了全球人工智能治理倡议(以下简称“倡议”),意义重大。人工智能的兴起引发了新一轮科技革命的浪潮,因此,人工智能急需治理,治理人工智能的重要性在于平衡安全与发展的重要性。

  人工智能治理强调以人为本,科技向善。如何让科技以人为本是使用人工智能的重点问题。今天,整个国际社会惊诧于人工智能新一轮热潮的迸发。在这种情况下,提出倡议可以使其更好服务全人类。倡议内容简短、语言简洁、问题切中要害,尤其是倡议中提到了一些风险问题,包括智能鸿沟的问题、全球的人工智能协调发展、新的南北差距等等。倡议用简洁明了的语言概括了当今全球人工智能发展面临的问题,几乎涵盖了人工智能治理的各个领域,涵盖了国际社会的各个核心关切,使国际社会可以更加具有前瞻性地去思考人工智能的未来发展道路。

  作为一个人工智能发展的大国,中国提出倡议本身体现出一种大国的责任和担当,也是一种风向标。倡议的提出,给人工智能未来的发展注入强劲的力量,吹响了人工智能改革的号角,为人工智能向善的未来指明了方向。

  人工智能的以人为本离不开风险防范。风险防范关乎安全。因此,我们要从更广义的角度去理解,一方面是这个技术本身是不是安全,另一方面是理解人工智能本身所存在的伦理道德问题。同时,我们需要看到不确定性。谈论人工智能主要聚焦在其是否能够带来技术变革,但变革的走向还存在不确定性。如果风险不能够预判的话,意味着风险高,治理过程会复杂化。

  人工智能治理需要跨学科、跨领域。未来,在防范风险方面,需要细化风险,比如说技术层面会有哪些风险,应用层面会带来哪些风险,然后规则层面会有哪些风险,甚至整个发展进程可能总体会对社会带来一些什么样的风险,人工智能也是同样的。如果要细化风险,各个层面其实都会有各种各样的安全问题,这就要求综合治理。

  不管是战略界、政策界、学术界、企业界,只有各行各业汇集到一起,才可能对人工智能发展的动态与未来的发展的方向有一个相对比较接近于客观现实的判断,为人工智能治理提出更好的决策。为了进一步规范人工智能治理,各个部门应当积极探索,包括深层次人工智能的管理规范,尤其是一些内容方面的,去维护人工智能安全,降低人工智能可能带来的一些安全风险,做一些人工智能的标准制定、技术评估。

  人工智能是新兴事物,各个部门需要保持开放、保持弹性,这对于完善人工智能治理十分重要。此外,对于机制的反馈也是十分重要的。我们需要在摸索中前进。

  就国际维度而言,国际社会也做出了不少努力。在联合国框架里,我们也对人工智能的发展发表了自己的主张,我们也会去配合或者积极响应国际上的一些议程,包括参加英国主办的人工智能峰会。在国际上,我们整体表现出一种开放的、积极的状态。

  今后,随着数字化的发展,不平等现象也会出现,南北差距会进一步加大,不利于人工智能向善。人工智能不应该赢者通吃,强者恒强,只有共同发展才能够造福国际社会、造福人类。对于广大发展中国家来说,为了争取主动权,应当主动去认识人工智能、迎接科技革命,只有这样才能够应对全球人工智能领域带来的挑战。

  中国长期作为发展中国家,在这个过程中,对于科技的变革,我们感同身受,我们也见证了历史发展的规律,只有行稳致远方可合理、健康运用人工智能。 从某种意义上来说,对于人工智能的未来,我们要考虑的就是全球的安全和稳定,即如何用人工智能推动全球安全稳定,这是世界各国需要做的功课。

  中国社会科学网 李想/整理

  (作者系中国现代国际关系研究院科技与网络安全研究所执行所长、研究员)

Li Yan: AI governance lies in balancing security and development
2024-05-11 Author: Li Yan Source: China Social Sciences Network

In October 2023, President Xi Jinping proposed the Global AI Governance Initiative (hereinafter referred to as the "Initiative") at the third Belt and Road Forum for International Cooperation, which is of great significance. The rise of AI has triggered a new wave of technological revolution. Therefore, AI is in urgent need of governance. The importance of AI governance lies in balancing security and development.

AI governance emphasizes people-oriented and technology for good. How to make technology people-oriented is the key issue in the use of AI. Today, the entire international community is surprised by the outbreak of a new wave of AI enthusiasm. In this case, proposing an initiative can make it better serve all mankind. The content of the initiative is short, the language is concise, and the issues are to the point. In particular, the initiative mentions some risk issues, including the problem of the intelligence gap, the coordinated development of global AI, the new North-South gap, and so on. The initiative summarizes the problems facing the development of global AI in concise and clear language, covering almost all areas of AI governance and all core concerns of the international community, so that the international community can think more forward-lookingly about the future development path of AI.

As a major country in the development of artificial intelligence, China's proposal itself reflects the responsibility and commitment of a major country, and is also a weather vane. The proposal has injected strong power into the future development of artificial intelligence, sounded the clarion call for artificial intelligence reform, and pointed out the direction for the future of artificial intelligence to be good.

The people-oriented nature of artificial intelligence is inseparable from risk prevention. Risk prevention is related to safety. Therefore, we need to understand from a broader perspective, on the one hand, whether the technology itself is safe, and on the other hand, understand the ethical and moral issues of artificial intelligence itself. At the same time, we need to see uncertainty. Talking about artificial intelligence mainly focuses on whether it can bring about technological change, but there is still uncertainty about the direction of change. If risks cannot be predicted, it means that the risk is high and the governance process will be complicated.

Artificial intelligence governance requires interdisciplinary and cross-domain. In the future, in terms of risk prevention, risks need to be refined, such as what risks will there be at the technical level, what risks will be brought at the application level, and what risks will be brought at the rule level, and even what kind of risks the entire development process may bring to society as a whole, and the same is true for artificial intelligence. If risks are to be refined, there will actually be various security issues at all levels, which requires comprehensive governance.

No matter whether it is the strategic community, policy community, academic community, or business community, only when all walks of life come together can we have a relatively objective and realistic judgment on the dynamics of AI development and the future direction of development, and make better decisions for AI governance. In order to further standardize AI governance, various departments should actively explore, including deep-level AI management norms, especially some content aspects, to maintain AI security, reduce some security risks that AI may bring, and make some AI standard formulation and technical evaluation.

Artificial intelligence is an emerging thing, and various departments need to remain open and flexible, which is very important for improving AI governance. In addition, feedback on the mechanism is also very important. We need to move forward in exploration.

In terms of the international dimension, the international community has also made a lot of efforts. In the framework of the United Nations, we have also expressed our own views on the development of AI, and we will also cooperate with or actively respond to some international agendas, including participating in the AI ​​Summit hosted by the United Kingdom. Internationally, we have shown an open and positive state as a whole.

In the future, with the development of digitalization, inequality will also appear, and the gap between the North and the South will further widen, which is not conducive to the good of AI. Artificial intelligence should not be a winner-takes-all game, and the strong should always be strong. Only common development can benefit the international community and mankind. For the vast majority of developing countries, in order to gain the initiative, they should take the initiative to understand artificial intelligence and welcome the scientific and technological revolution. Only in this way can they cope with the challenges brought by the global artificial intelligence field.

China has been a developing country for a long time. In this process, we feel the same about the changes in science and technology. We have also witnessed the laws of historical development. Only by moving steadily and far can we use artificial intelligence reasonably and healthily. In a sense, for the future of artificial intelligence, what we need to consider is global security and stability, that is, how to use artificial intelligence to promote global security and stability. This is the homework that countries around the world need to do.

China Social Sciences Network Li Xiang/Compilation

(The author is the executive director and researcher of the Institute of Science and Technology and Cyber ​​Security of the China Institute of Contemporary International Relations)

 


 【聚焦学理中国】《全球人工智能治理倡议》中的中国智慧
2024-01-23 作者:鲁传颖 来源:中国社会科学网

2023年10月18日,中国政府发布《全球人工智能治理倡议》,围绕人工智能发展、安全、治理三方面系统阐述了人工智能治理中国方案。《全球人工智能治理倡议》既是在人工智能技术快速发展,机遇与风险快速涌现的关键时期贡献的中国智慧,同样也是在人工智能治理的规则建构被西方国家垄断的背景下代表广大发展中国家发出的呼吁,将会对全球人工智能发展与治理、人类命运共同体的愿景落实起到积极的作用。

  一、《全球人工智能治理倡议》设计了一条不同于西方国家的治理思路

  人工智能由于具有战略性、全局性、不确定性等特征,已经成为影响国家发展和安全的重要力量,世界各国纷纷出台人工智能国家战略,并且开始在人工智能治理的国际规则制定领域展开竞争与合作,相关的倡议、愿景和规划如雨后春笋般涌现。在此背景下,中国的《全球人工智能治理倡议》在以下三个方面展现出了不同于西方国家的独特性和优越性。

  首先,中国强调人工智能发展必须坚持“以人为本”理念。这种理念源于中国古代“民本”思想,强调国家和社会应服务于人民,同时也吸收了西方关于人权的讨论,批判了其中过于强调工具性和武器化的倾向,提倡更为和谐、平衡的人类关系视角。这种“人本”思想在当今人工智能时代得到了新的内涵和重要性,即人工智能技术的发展不仅需要尊重和维护个人权利,更应该以增进人类共同福祉为目标,以保障社会安全、尊重人类权益为前提,确保人工智能始终朝着有利于人类文明进步的方向发展,并进一步助力气候、生物等可持续发展议题的落实。从这个意义上来讲,倡议不仅为人工智能技术的道德规范和发展方向提供了框架,更提出了更为深刻的伦理立意,倡导构建一个平衡各方利益、共同应对人类挑战的全球人工智能治理新模式。

  其次,中国强调发展人工智能发展需要跨越“小圈子”,反对特定国家将全球治理机制工具化、私物化。《全球人工智能治理倡议》强调应坚持相互尊重、平等互利的原则,各国无论大小、强弱,无论社会制度如何,都有平等发展和利用人工智能的权利。这是对西方国家日益倡导的“小圈子”治理模式的一次修正和突破,强调人工智能的发展必须惠及各国人民,而不应该成为特定国家巩固科技霸权、攫取私利的工具。相比于此,美国、英国等西方国家虽然积极开展人工智能国际合作,但不论是军事防务领域的人工智能防务伙伴关系(AI PfD)、美英澳三国集团(AUKUS),还是产业与发展领域的美欧贸易和技术委员会(TTC)以及美印人工智能倡议(USIAI),抑或是伦理与价值方面的自由在线联盟人工智能与人权工作组(T-FAIR)等机制,都有着极强的排他性和竞争性。如果不能破除这种思维,建立真正以发展和安全为导向的人工智能全球治理机制,则会人为放大人工智能发展中的风险,阻碍人工智能的全球治理进程。

  最后,中国站在了广大发展中国家的立场上,提出需要增强发展中国家在人工智能全球治理中的代表性和发言权,让广大发展中国家更加平等地参与其中。从现实来看,发展中国家在人工智能的全球治理和国际规则制定上存在参与程度不深入、参与范围较狭隘的问题。这一方面与广大发展中国家的人工智能发展还处于落后阶段、参与国际规则的动力和意识不强有关,另一方面,发达国家对治理机制和国际规则的垄断则进一步强化了发展中国家无法参与国际治理的困境。在此背景下,中国的《全球人工智能治理倡议》提出了各国人工智能发展与治理的“权利平等”“机会平等”“规则平等”的原则,并倡导需要开展面向发展中国家的国际合作与援助,不断弥合智能鸿沟和治理能力差距。这将从能力、意愿与平台开放度两方面破解发展中国家在人工智能全球治理中的边缘化难题,通过兼顾西方与非西方的国家立场和声音建构公平合理的国际规则。

  二、《全球人工智能治理倡议》提供了一种切实可行的人工智能治理方案

  事实上,中国发布《全球人工智能治理倡议》的目的并非是为了“立异”,更是为了“求同”。《全球人工智能治理倡议》在面向现实、面向未来的基础上,不仅对当前由美欧等西方国家主导的人工智能全球治理方案提供了一定的修正和补充,也进一步为落实当前人工智能治理的已有共识提供了实践方案。因此,《全球人工智能治理倡议》不仅是一份愿景,也是一份行动指南。

  第一,《全球人工智能治理倡议》在人工智能全生命周期治理中提供给了关键指导。由于人工智能风险蕴含在其技术设计、研发、应用等不同的发展阶段中,且环环相扣,相互叠加。因此,必须要以全生命周期的视角进行风险识别和治理方案的设计。在《全球人工智能治理倡议》中,既提出了在概念化设计阶段的“伦理先行”原则;也提出研发阶段的数据获取、算法设计、技术开发、产品研发等程序中要坚持公平和非歧视的原则;还强调了在部署和应用阶段要坚持和落实“智能向善”的宗旨;并进一步要求要建立跨阶段、快速响应、敏捷高效的风险等级测试评估体系。通过全生命周期的治理机制设计,确保人工智能始终处于人类控制之下,打造可审核、可监督、可追溯、可信赖的人工智能技术。

  第二,《全球人工智能治理倡议》对全球人工智能治理的共识性原则提出了落实建议。比如说,中国在多边治理的原则下明确提出要继续坚持联合国发挥主导性作用,支持在联合国框架下讨论成立国际人工智能治理机构,协调国际人工智能发展、安全与治理重大问题;还比如说,在人工智能伦理方面,中国则明确提出建立不仅要在事前建立并完善人工智能伦理准则、规范形成人工智能伦理指南,明确人工智能相关主体的责任和权力边界;还要建立事中和事后的科技伦理问责制度、审查和监管制度;更在充分尊重并保障各群体合法权益的基础上及时回应国内和国际相关伦理关切。由此可见,中国的这一倡议展现了深远的战略视野和务实的行动方案,不仅回应了当前国际社会对人工智能发展的普遍关切,也为全球人工智能治理提供了新实施路径。通过这样的努力,中国正积极参与全球治理的创新,推动构建更加公正、合理的国际人工智能治理新秩序。

  第三,《全球人工智能治理倡议》将安全有序发展的中国经验介绍给了世界。对于人工智能治理而言,安全是发展的前提,如果没有安全,发展取得的成果也会得而复失,而发展又是安全的保障,如果不能以动态的视角跟进技术发展的步伐,也就无法实现有效的治理。因此,从中国的视角来看,人工智能治理的目的要在始终维持发展进程的基础上,通过相关的规则建构,确保人工智能技术安全,增进人类的共同福祉。为了落实这一点,中国在《全球人工智能治理倡议》中设计了广泛参与、协商一致、循序渐进的原则,将治理融入到发展的进程中,同时也提出积极发展用于人工智能治理的相关技术开发与应用,使技术发展赋能安全治理。由此,中国设计了一套在发展与安全,变革与稳定等多重维度之间寻找合理的平衡点的具体思路。

  三、《全球人工智能治理倡议》将对全球治理的未来产生重要影响

  由于理念的先进性和现实的可操作性,中国《全球人工智能治理倡议》不仅为国际社会提供了一个具体、可行的行动纲领,也展示了中国在全球治理中所持的开放态度和建设性立场。这一倡议将进一步推动全球人工智能治理的进程,并加速实现中国关于“共建人类命运共同体”的宏伟愿景。

  从人工智能治理的全球进程来看,中国《全球人工智能治理倡议》提供了一个非约束性的开放框架。这一倡议只规定了底线,提供了合作的平台,因此为破除当前人工智能全球治理的僵局提供了更多可能性。当前,全球人工智能治理面临诸多挑战:在国际安全上面临着人工智能军事化、自主武器的使用与规范的问题;在技术安全上面临着标准制定困难、滞后、约束性弱等问题;在技术创新方面出现了各国的信任感不足,以战略竞争取代技术的开放合作的问题;在技术应用方面出现了数字鸿沟和不平等发展问题……这些难题阻碍人工智能治理进程的推进。在此背景下,中国的倡议通过提供一个开放和包容的框架,鼓励各国根据自己的国情和需求进行选择和调整,为各国之间就人工智能治理问题的合作提供了新机遇。

  而从人类命运共同体理念的落实来看,《全球人工智能治理倡议》将与“三大全球倡议”一起,共同为建设国际新秩序、落实人类命运共同体的愿景贡献力量。面对世界之变、时代之变、历史之变,中国通过“三大全球倡议”为全球人民普遍关注的发展、安全与合作等关键问题提供了中国的解释和方案的同时,也并未止步于传统议题的探讨。中国敏锐地洞察到新兴技术给全球治理带来的新问题和挑战,通过提出《全球人工智能治理倡议》展现了积极承担国际责任的姿态。中国希望通过自身的努力,促进技术发展、安全与治理的和谐统一。在此基础上,各国将共同分享人工智能等新兴技术的发展成果,实现数字鸿沟的跨越。而在这一系列倡议的背后是不断发出的中国之声和不断涌现的中国智慧。中国将在传统和新兴领域持续发力,实现人类命运共同体的美好愿景。

  (作者系上海国际问题研究院研究员、网络空间国际治理研究中心秘书长)

 

【Focus on Academic China】Chinese Wisdom in the Global Initiative on Artificial Intelligence Governance
2024-01-23 Author: Lu Chuanying Source: China Social Sciences Network

On October 18, 2023, the Chinese government issued the Global Initiative on Artificial Intelligence Governance, which systematically expounded the Chinese solution for artificial intelligence governance in three aspects: artificial intelligence development, security, and governance. The Global Initiative on Artificial Intelligence Governance is not only the Chinese wisdom contributed in a critical period when artificial intelligence technology is developing rapidly and opportunities and risks are emerging rapidly, but also an appeal on behalf of the vast number of developing countries in the context of the monopoly of the rule construction of artificial intelligence governance by Western countries. It will play a positive role in the development and governance of global artificial intelligence and the implementation of the vision of a community with a shared future for mankind.

I. The Global Initiative on Artificial Intelligence Governance has designed a governance approach different from that of Western countries

Due to its strategic, global, and uncertain characteristics, artificial intelligence has become an important force affecting national development and security. Countries around the world have introduced national artificial intelligence strategies and have begun to compete and cooperate in the field of international rule-making for artificial intelligence governance. Related initiatives, visions, and plans have sprung up like mushrooms after rain. Against this background, China's "Global AI Governance Initiative" shows its uniqueness and superiority in the following three aspects, which are different from those of Western countries.

First, China emphasizes that the development of AI must adhere to the "people-oriented" concept. This concept originates from the ancient Chinese "people-oriented" thought, emphasizing that the state and society should serve the people. At the same time, it also absorbs the Western discussion on human rights, criticizes the tendency of over-emphasis on instrumentality and weaponization, and advocates a more harmonious and balanced perspective on human relations. This "people-oriented" thought has gained new connotations and importance in today's era of artificial intelligence, that is, the development of artificial intelligence technology not only needs to respect and safeguard individual rights, but also should be aimed at promoting the common welfare of mankind, with the premise of ensuring social security and respecting human rights, ensuring that artificial intelligence always develops in a direction that is conducive to the progress of human civilization, and further helps the implementation of sustainable development issues such as climate and biology. In this sense, the initiative not only provides a framework for the ethical norms and development direction of artificial intelligence technology, but also puts forward a deeper ethical intention, advocating the construction of a new model of global AI governance that balances the interests of all parties and jointly responds to human challenges.

Second, China emphasizes that the development of artificial intelligence needs to transcend "small circles" and opposes the instrumentalization and privatization of global governance mechanisms by specific countries. The Global AI Governance Initiative emphasizes the principle of mutual respect, equality and mutual benefit. All countries, regardless of size, strength or social system, have the right to develop and use AI equally. This is a correction and breakthrough to the "small circle" governance model increasingly advocated by Western countries, emphasizing that the development of AI must benefit the people of all countries, and should not become a tool for specific countries to consolidate their technological hegemony and gain private interests. In contrast, although Western countries such as the United States and the United Kingdom are actively engaged in international cooperation in AI, whether it is the AI ​​Defense Partnership (AI PfD) in the field of military defense, the US-UK-Australia Group (AUKUS), the US-EU Trade and Technology Committee (TTC) in the field of industry and development, the US-India Artificial Intelligence Initiative (USIAI), or the Free Online Alliance Artificial Intelligence and Human Rights Working Group (T-FAIR) in the field of ethics and values, all have extremely strong exclusivity and competitiveness. If we cannot break this kind of thinking and establish a truly development- and security-oriented global AI governance mechanism, we will artificially magnify the risks in the development of AI and hinder the global governance process of AI.

Finally, China stood on the standpoint of the vast majority of developing countries and proposed that the representation and voice of developing countries in the global governance of artificial intelligence should be enhanced, so that the vast majority of developing countries can participate more equally. From a realistic perspective, developing countries have problems with their lack of participation and narrow scope in the global governance of artificial intelligence and the formulation of international rules. On the one hand, this is related to the fact that the development of artificial intelligence in the vast majority of developing countries is still in a backward stage and their motivation and awareness of participating in international rules are not strong. On the other hand, the monopoly of developed countries on governance mechanisms and international rules has further strengthened the dilemma of developing countries' inability to participate in international governance. Against this background, China's "Global Initiative on Artificial Intelligence Governance" proposed the principles of "equal rights", "equal opportunities" and "equal rules" for the development and governance of artificial intelligence in various countries, and advocated the need to carry out international cooperation and assistance for developing countries to continuously bridge the intelligence gap and governance capacity gap. This will solve the marginalization problem of developing countries in the global governance of artificial intelligence from the two aspects of capacity, willingness and platform openness, and construct fair and reasonable international rules by taking into account the positions and voices of Western and non-Western countries.

II. The "Global Initiative on Artificial Intelligence Governance" provides a practical and feasible solution for artificial intelligence governance

In fact, the purpose of China's release of the "Global Initiative on Artificial Intelligence Governance" is not to "innovate", but to "seek common ground". The Global Initiative on AI Governance, based on the reality and the future, not only provides certain corrections and supplements to the current AI global governance plan led by Western countries such as the United States and Europe, but also further provides a practical plan for implementing the existing consensus on AI governance. Therefore, the Global Initiative on AI Governance is not only a vision, but also a guide to action.

First, the Global Initiative on AI Governance provides key guidance in the governance of AI throughout its life cycle. Since AI risks are contained in different development stages such as its technical design, R&D, and application, and are closely linked and superimposed on each other. Therefore, it is necessary to identify risks and design governance plans from a full life cycle perspective. In the Global Initiative on AI Governance, the principle of "ethics first" is proposed in the conceptual design stage; the principle of fairness and non-discrimination should be adhered to in the procedures of data acquisition, algorithm design, technology development, and product development in the R&D stage; it also emphasizes the need to adhere to and implement the purpose of "intelligent for good" in the deployment and application stages; and further requires the establishment of a cross-stage, rapid response, agile and efficient risk level testing and evaluation system. Through the design of a governance mechanism throughout the life cycle, we can ensure that artificial intelligence is always under human control and create auditable, supervised, traceable and trustworthy artificial intelligence technology.

Second, the Global AI Governance Initiative proposes implementation suggestions for the consensus principle of global AI governance. For example, under the principle of multilateral governance, China has clearly proposed to continue to insist on the United Nations playing a leading role, support the discussion of the establishment of an international AI governance organization under the United Nations framework, and coordinate major issues of international AI development, security and governance; for example, in terms of AI ethics, China has clearly proposed to establish not only to establish and improve AI ethics standards in advance, to standardize the formation of AI ethics guidelines, and to clarify the responsibilities and power boundaries of AI-related entities; but also to establish a science and technology ethics accountability system, review and supervision system during and after the event; and to respond to relevant domestic and international ethical concerns in a timely manner on the basis of fully respecting and protecting the legitimate rights and interests of all groups. It can be seen that China's initiative has demonstrated a far-reaching strategic vision and a pragmatic action plan, which not only responds to the current international community's general concerns about the development of AI, but also provides a new implementation path for global AI governance. Through such efforts, China is actively participating in the innovation of global governance and promoting the construction of a more fair and reasonable new order of international AI governance.

Third, the Global Initiative on AI Governance introduces China's experience of safe and orderly development to the world. For AI governance, security is the premise of development. Without security, the achievements of development will be lost, and development is the guarantee of security. If we cannot keep up with the pace of technological development from a dynamic perspective, we will not be able to achieve effective governance. Therefore, from China's perspective, the purpose of AI governance is to ensure the security of AI technology and enhance the common welfare of mankind through the construction of relevant rules on the basis of always maintaining the development process. In order to implement this, China has designed the principles of broad participation, consensus, and gradual progress in the Global Initiative on AI Governance, integrating governance into the process of development. At the same time, it also proposes to actively develop relevant technology development and application for AI governance, so that technological development can empower safe governance. As a result, China has designed a set of specific ideas to find a reasonable balance between multiple dimensions such as development and security, change and stability.

Third, the Global Initiative on AI Governance will have an important impact on the future of global governance

Due to the advanced concept and practical operability, China's Global Initiative on AI Governance not only provides a specific and feasible action program for the international community, but also demonstrates China's open attitude and constructive position in global governance. This initiative will further promote the process of global AI governance and accelerate the realization of China's grand vision of "building a community with a shared future for mankind".

From the perspective of the global process of AI governance, China's "Global AI Governance Initiative" provides a non-binding open framework. This initiative only sets the bottom line and provides a platform for cooperation, thus providing more possibilities for breaking the current deadlock in global AI governance. At present, global AI governance faces many challenges: in international security, it faces the problems of AI militarization, the use and regulation of autonomous weapons; in technical security, it faces the problems of difficulty, lag and weak binding in standard setting; in technological innovation, there is a lack of trust among countries, and strategic competition replaces open cooperation in technology; in technology application, there are problems of digital divide and unequal development... These problems hinder the advancement of AI governance. In this context, China's initiative provides an open and inclusive framework to encourage countries to make choices and adjustments according to their national conditions and needs, and provides new opportunities for cooperation between countries on AI governance issues.

From the perspective of the implementation of the concept of a community with a shared future for mankind, the "Global AI Governance Initiative" will work together with the "three major global initiatives" to contribute to the construction of a new international order and the implementation of the vision of a community with a shared future for mankind. In the face of changes in the world, the times, and history, China has provided China's interpretation and solutions for key issues such as development, security, and cooperation that are of common concern to the people of the world through the "Three Global Initiatives", and has not stopped at the discussion of traditional issues. China has keenly perceived the new problems and challenges that emerging technologies bring to global governance, and has demonstrated its attitude of actively assuming international responsibilities by proposing the "Global Initiative on Artificial Intelligence Governance". China hopes to promote the harmonious unity of technological development, security, and governance through its own efforts. On this basis, countries will share the development achievements of emerging technologies such as artificial intelligence and achieve the bridging of the digital divide. Behind this series of initiatives is the continuous voice of China and the continuous emergence of Chinese wisdom. China will continue to exert its strength in traditional and emerging fields to realize the beautiful vision of a community with a shared future for mankind.

(The author is a researcher at the Shanghai Institutes for International Studies and Secretary-General of the Center for International Governance of Cyberspace) 


【聚焦学理中国】《全球人工智能治理倡议》为数字公平贡献中国方案
2024-02-01 作者:全燕 来源:中国社会科学网

为规范全世界人工智能的行为准则,为未来世界的人工智能提出治理的规划、降低人工智能对人类社会带来的风险,2023年10月18日,中国向世界提出了《全球人工智能治理倡议》(以下简称《倡议》),为推进全球人工智能治理提出了宝贵方案。中国社会科学网约请广东外语外贸大学新闻与传播学院教授、“云山杰出学者”全燕,对《倡议》进行解读。

  中国社会科学网:《倡议》如何从国际层面规范人工智能运用?

  全燕:全球人工智能治理是全球治理的一个重要的内涵。全球治理的一个总体目标是向全球提供公共产品,并且要避免全球性的公共危害。今天,全球公共产品日趋依赖于人工智能的支持,人工智能技术本身在某种程度上也可以被视为一种全球公共产品。因此,全球人工智能治理可定义为由各国政府、国际和政府间组织,通过制定规则与程序来规范人工智能系统的开发和部署。

  倡议提出了与人工智能技术相关的具有规范意义的问题,制定了使人工智能技术开发产生正向社会效应所必须遵循的规范。通常情况下,人工智能技术的开发和部署都是在跨国和跨地区范围内进行的,很少考虑传统辖区的限制,而全球人工智能治理考虑的是针对人工智能的全球性规范政策的制定与实施。倡议内容涉及管理和引导人工智能的发展,以确保人工智能在全球范围内的应用是安全公正且符合伦理标准的。

  首先,需要全面理解应该受到治理的主体。如何治理参与普遍开发和部署人工智能技术的全球主体是关键问题。倡议强调,发展人工智能应坚持“智能向善”的宗旨,遵守适用的国际法,符合和平、发展、公平、正义、民主、自由的全人类共同价值。我国贡献的治理方法论和元理论,通过一系列的期望提供理论框架,为更加具体的解决方案设定了原则和规范的界限。

  其次,要坚持公平性和非歧视性的原则。决策算法可能会反映和加剧现有的不公正,是算法偏见的一个集中体现。倡议中提到,要避免在数据获取、算法设计、技术开发、产品研发与应用过程中,产生针对不同或特定民族信仰、国别、性别等偏见和歧视。

  此外,人工智能的治理还应该注意到利润分配。技术利润是私有化的,而技术的负面效应是社会化的。人工智能技术可以为开发者和持有者带来利润,但它可能产生的有害影响以及风险在很大程度上是由社会共担。

  中国社会科学网:《倡议》对实现数字平等的意义在哪里?

  全燕:实现数字平等的关键因素在于扩大透明度。为了确保人工智能在部署时能够遵循监管的框架,还需要更多的工具来支持治理。可解释的机器学习对于适当信任和有效管理新一代人工智能机器至关重要。无论国家和地区的差异多大,人工智能治理都需要得到验证工具和参与激励机制的支持。验证工具可以帮助国内、国际参与者之间建立对技术的必要信任,包括算法设计和监管机构用于评估人工智能实施的方法,例如数据科学方法、算法设计、硬件等。如果我们缺乏验证方法,对人工智能的有效治理就将面临无法深入的困境。

  有效的全球人工智能治理需要各国、各地区的参与,创建共同宪章或者公约。这就需要实施共享技术数据和专业知识等激励措施,促使多国、多边、多方能够参与进来。如果世界上大部分地区没有参与制定和发展这些宪章,全球人工智能治理就无法得到广泛支持,也就没有实际意义。多方主体的参与和共同宪章的创建将有助于促进有效治理,有效减少数字的不平等。

  当前,世界各国各地区都颁布了相应的数据保护条例,欧盟颁布《通用数据保护条例》和人工智能伦理原则,美国签署了阿西洛马人工智能23条原则,中国也陆续出台包括数据隐私保护、网络安全等级保护等法律法规。虽然各国都出台了相应治理方案,但是国家和区域间的差异较为明显,原则和方法不尽相同。只有就人工智能治理相关问题建立更大的对话机制,让各国都能够清晰地找到共同利益和争议点所在,并以此为基准,全球治理才能够实现实质性推进。因此,在构建起更广泛的全球框架之前,国际社会有必要建立特定的人工智能治理机制和标准,如规范致命自主武器系统、医疗保健、交通和通用人工智能等领域的应用标准等。在促进公正、平等的基础上,利用更加精准的法规和更恰当的治理机制,包括用于认证、解释和审计算法系统的方法等,都将有助于确保对人工智能的伦理关系能得到回应。

  中国社会科学网:在您看来,《倡议》如何影响人类文明发展?

  全燕:《倡议》对人类文明发展具有重要意义。全球人工智能治理倡议为国际社会规范使用人工智能提供了重要参考。

  首先为人类共同福祉而努力。《倡议》强调了以人为本的理念,也就是将人工智能的发展目标定位在增进人类共同的福祉,而这种理念有助于强调各国共同的价值目标,为国际合作提供动力。

  其次促进国际合作。《倡议》提倡各国在人工智能领域的信息交流和技术合作,这种国际合作有助于加深各国之间的相互了解,同时也能够促进在技术研发和标准制定等方面的共同努力。

  再次促进平等互利。《倡议》强调相互尊重、平等互利的原则,鼓励各国能够平等地参与人工智能技术的发展,这将有助于减少不同国家之间不平等的现象。尤其是缩小技术发达国家和不发达国家或者欠发达国家之间的技术鸿沟,促进共同繁荣。

  再其次有利于文明互鉴。《倡议》重点强调智能向善的宗旨,其实就是遵守适用的国际法和全人类共同价值观,这也将有助于促进不同文明之间的相互尊重、学习和交流。

  最后促进伦理与公平。《倡议》提出建立人工智能的伦理准则和规范,注重公平和非歧视性,这些将有助于在人工智能发展中建立起共同的伦理价值观,促进各国在技术应用中避免歧视性行为。

  当前,智能文明正在影响着人类文明的发展方向。《倡议》中提到了“智能向善”,我将其具体化为“算法向善”,也就是如何促进在算法造福人类或者说人工智能造福人类的这样宗旨面前,控制由技术带给我们的风险。人工智能的确改变了我们的文明形态,但从另一个方面讲,利用好人工智能,就能够更好地促进人类文明新形态的建设。

  《倡议》为不同国家和文明之间搭建了一个共同的平台,通过合作与对话来促进共同的发展,共同应对人工智能技术的挑战,同时推动国际社会在人工智能治理方面形成一个共识,从而促进中国与世界各国的文明交流。

  中国社会科学网 李想/整理

[Focus on Academic China] "Global AI Governance Initiative" contributes Chinese solutions to digital equity
2024-02-01 Author: Quan Yan Source: China Social Sciences Network

In order to standardize the code of conduct of artificial intelligence in the world, propose a governance plan for artificial intelligence in the future world, and reduce the risks that artificial intelligence brings to human society, on October 18, 2023, China proposed the "Global AI Governance Initiative" (hereinafter referred to as the "Initiative") to the world, and put forward a valuable plan for promoting global artificial intelligence governance. China Social Sciences Network invited Quan Yan, professor of the School of Journalism and Communication of Guangdong University of Foreign Studies and "Yunshan Outstanding Scholar", to interpret the "Initiative".

China Social Sciences Network: How does the "Initiative" regulate the use of artificial intelligence at the international level?

Quan Yan: Global AI governance is an important connotation of global governance. One of the overall goals of global governance is to provide public goods to the world and avoid global public hazards. Today, global public goods are increasingly dependent on the support of artificial intelligence, and artificial intelligence technology itself can also be regarded as a global public product to some extent. Therefore, global AI governance can be defined as the development and deployment of artificial intelligence systems by governments, international and intergovernmental organizations through the formulation of rules and procedures.

The initiative raises normative issues related to AI technology and sets out the norms that must be followed to make the development of AI technology produce positive social effects. Generally, the development and deployment of AI technology are carried out on a transnational and transregional scale, rarely considering the limitations of traditional jurisdictions, while global AI governance considers the formulation and implementation of global normative policies for AI. The content of the initiative involves managing and guiding the development of AI to ensure that the application of AI on a global scale is safe, fair and ethical.

First, it is necessary to fully understand the subjects that should be governed. How to govern global subjects involved in the universal development and deployment of AI technology is a key issue. The initiative emphasizes that the development of AI should adhere to the purpose of "intelligence for good", comply with applicable international law, and comply with the common values ​​of peace, development, fairness, justice, democracy and freedom for all mankind. The governance methodology and meta-theory contributed by my country provide a theoretical framework through a series of expectations, setting the boundaries of principles and norms for more specific solutions.

Second, we must adhere to the principles of fairness and non-discrimination. Decision-making algorithms may reflect and exacerbate existing injustices, which is a concentrated manifestation of algorithmic bias. The initiative mentioned that it is necessary to avoid bias and discrimination against different or specific ethnic beliefs, nationalities, genders, etc. in the process of data acquisition, algorithm design, technology development, product development and application.

In addition, the governance of artificial intelligence should also pay attention to profit distribution. Technology profits are privatized, while the negative effects of technology are socialized. Artificial intelligence technology can bring profits to developers and holders, but the harmful effects and risks it may produce are largely shared by society.

China Social Sciences Network: What is the significance of the "Initiative" for achieving digital equality?

Quan Yan: The key factor in achieving digital equality is to expand transparency. In order to ensure that artificial intelligence can follow the regulatory framework when deployed, more tools are needed to support governance. Explainable machine learning is essential for proper trust and effective management of the new generation of artificial intelligence machines. Regardless of the differences between countries and regions, artificial intelligence governance needs to be supported by verification tools and incentive mechanisms for participation. Verification tools can help build the necessary trust in technology between domestic and international participants, including algorithm design and methods used by regulators to evaluate the implementation of artificial intelligence, such as data science methods, algorithm design, hardware, etc. If we lack verification methods, effective governance of artificial intelligence will face the dilemma of not being able to go deeper.

Effective global AI governance requires the participation of all countries and regions to create a common charter or convention. This requires the implementation of incentives such as sharing technical data and expertise to encourage multinational, multilateral and multi-party participation. If most parts of the world do not participate in the formulation and development of these charters, global AI governance will not be widely supported and will have no practical significance. The participation of multiple parties and the creation of a common charter will help promote effective governance and effectively reduce digital inequality.

Currently, countries and regions around the world have promulgated corresponding data protection regulations. The European Union has promulgated the General Data Protection Regulation and the Artificial Intelligence Ethics Principles. The United States has signed the Asilomar Artificial Intelligence 23 Principles. China has also successively issued laws and regulations including data privacy protection and cybersecurity level protection. Although all countries have issued corresponding governance plans, the differences between countries and regions are relatively obvious, and the principles and methods are not the same. Only by establishing a larger dialogue mechanism on issues related to AI governance, so that all countries can clearly find common interests and controversial points, and use this as a benchmark, can global governance achieve substantial progress. Therefore, before building a broader global framework, it is necessary for the international community to establish specific AI governance mechanisms and standards, such as application standards for lethal autonomous weapons systems, healthcare, transportation, and general AI. On the basis of promoting fairness and equality, using more precise regulations and more appropriate governance mechanisms, including methods for certifying, explaining, and auditing algorithmic systems, will help ensure that ethical relationships with AI are responded to.

China Social Sciences Network: In your opinion, how does the Initiative affect the development of human civilization?

Quan Yan: The Initiative is of great significance to the development of human civilization. The Global AI Governance Initiative provides an important reference for the international community to regulate the use of AI.

First, work for the common welfare of mankind. The Initiative emphasizes the concept of people-oriented, that is, positioning the development goal of AI to enhance the common welfare of mankind, and this concept helps to emphasize the common value goals of countries and provide impetus for international cooperation.

Second, promote international cooperation. The Initiative advocates information exchange and technical cooperation among countries in the field of AI. This international cooperation helps to deepen mutual understanding among countries and promote joint efforts in technology research and development and standard setting.

Third, promote equality and mutual benefit. The Initiative emphasizes the principles of mutual respect, equality and mutual benefit, and encourages all countries to participate equally in the development of artificial intelligence technology, which will help reduce inequality between different countries. In particular, it will narrow the technological gap between technologically developed countries and underdeveloped or underdeveloped countries, and promote common prosperity.

Secondly, it is conducive to mutual learning among civilizations. The Initiative emphasizes the purpose of intelligence for good, which is actually to abide by applicable international law and common values ​​of all mankind, which will also help promote mutual respect, learning and exchanges among different civilizations.

Finally, it promotes ethics and fairness. The Initiative proposes to establish ethical standards and norms for artificial intelligence, focusing on fairness and non-discrimination, which will help establish common ethical values ​​in the development of artificial intelligence and promote countries to avoid discriminatory behavior in the application of technology.

Currently, intelligent civilization is affecting the development direction of human civilization. The Initiative mentions "intelligent for good", which I will concretize as "algorithm for good", that is, how to promote the control of risks brought to us by technology in the face of the purpose of algorithms benefiting mankind or artificial intelligence benefiting mankind. Artificial intelligence has indeed changed our civilization, but on the other hand, making good use of artificial intelligence can better promote the construction of a new form of human civilization.

The Initiative builds a common platform for different countries and civilizations to promote common development through cooperation and dialogue, jointly respond to the challenges of artificial intelligence technology, and promote the international community to form a consensus on artificial intelligence governance, thereby promoting civilization exchanges between China and other countries in the world.

China Social Sciences Network Li Xiang/Compilation



全球人工智能治理倡议的世界意义
2024-06-19 作者:费艳颖 赵凯 来源:中国社会科学网-中国社会科学报

习近平总书记强调,让数字文明造福各国人民,推动构建人类命运共同体。2023年10月18日,习近平主席在第三届“一带一路”国际合作高峰论坛开幕式发表主旨演讲,提出“全球人工智能治理倡议”(以下简称“倡议”)。倡议着眼于数字文明时代的人工智能治理全球合作,标志着我国在人工智能领域积极践行人类命运共同体理念,落实全球发展倡议、全球安全倡议、全球文明倡议上做出了具体行动。全球人工智能治理倡议着力解决智能时代“人类文明向何处去?”这一历史之问,为数字文明时代的人工智能全球合作与有效治理贡献了建设性思路,为人工智能领域的国际讨论和规则制定提供了蓝本,发出了数字文明时代引领全球人工智能治理的“中国强音”。

  助力全球人工智能健康发展

  倡议为全球人工智能治理提供了理念引领,在统筹人工智能发展与安全下实现人工智能的全球协同共治,既为我国人工智能发展指明了方向,也为保障世界人工智能健康有序安全发展注入了磅礴的中国力量。其一,倡议突出“负责任”“以人为本”“智能向善”等理念,为全球人工智能治理提供价值指引。不同于个别国家极力构筑排他性“小圈子”,倡议秉持共同、综合、合作、可持续的安全观与共商共建共享的全球治理观,倡导以人为本、智能向善、多元参与、开放对话、共建共享、合作共赢,积极弥合智能鸿沟。倡议以更宽广的全球视野、更广泛的人工智能合作领域、更深层次的人工智能前瞻布局、更高水平的人工智能治理能力建设,为全球人工智能治理提供建设性解决思路。其二,统筹全球人工智能发展与安全两个关键点。倡议“坚持发展和安全并重的原则”,以敏捷治理和适度约束为原则,既克服了强力监管对于人工智能技术创新的不利影响,又避免了弱监管导致人工智能无序发展的乱象,实现人工智能技术发展与安全的平衡。其三,协调好人工智能国家治理和人工智能全球治理的关系。倡议充分反映了近年来中国在人工智能领域的立法、战略规划、伦理规范等成熟治理经验,还吸收了包括广大发展中国家在内的国际社会多边各方在人工智能治理领域的优秀成果。倡议坚持统筹国内和国际两个大局、着眼本国和世界,主张提升全球人工智能治理能力,既要构建开放、健康、安全的国内人工智能负责任创新生态,也要积极帮助其他发展中国家跨越智能鸿沟,改变全球人工智能发展领域极度不均衡的现状。

  化解全球人工智能安全风险

  倡议以科学的思维和方法,从中华文明的文化自信高度关注异质文明间的包容与互鉴,为推动全球人工智能健康发展凝聚广泛共识,为防范和化解全球人工智能安全风险贡献中国智慧。其一,倡议运用“七大思维”,为全球人工智能治理提供了科学的方法论指导。倡议指出,在人工智能的数据获取、算法设计、技术开发、产品研发与应用的全生命周期实现协同与系统治理;全局性思考人工智能治理问题,并将其上升到攸关全人类命运的战略高度;深刻洞察当前人工智能技术快速发展的历史方位,把握人工智能技术变革历史机遇期;主张人工智能发展和安全并重,国内和国际兼顾;创新治理理念,以敏捷治理助推“智能向善”,积极发展人工智能治理新技术;探索人工智能治理法治化路径;强调发展人工智能以尊重他国主权为底线,确保人工智能朝着有利于人类文明进步的方向发展。其二,倡议贯彻“六个坚持”,为全球人工智能治理提供了根本价值遵循。坚持人民至上,以“以人为本”理念促进人工智能技术辅助人类、增利人类、关怀人类;坚持自信自立,鼓励各主体协力共同促进人工智能治理,展现自信自立中国形象;坚持守正创新,紧跟人工智能前沿科技发展步伐,深化技术发展规律性认识;坚持问题导向,聚焦防范人工智能偏见和歧视、弥合全球智能鸿沟和治理能力差距等问题;坚持系统观念,既立足于全球人工智能治理需求,又着眼于我国人工智能治理实践,既观照当下,又瞄准未来,既谋求人工智能治理的统一性,又尊重各国政策和实践的多元性;坚持胸怀天下,兼具世界性与中国性,反对人为制造人工智能发展壁垒,主张以对话合作促进人工智能造福人类,彰显宽广的全球视野与世界胸怀。其三,倡议突出中华文明的创新性、包容性和和平性的独特底色,从文明高度为全球人工智能治理实践赋予中国智慧。倡议坚持“以人为本”、尊重他国主权、反对以意识形态划线或构建排他性集团、尊重各国人工智能政策和实践差异等核心主张,处处浸润着“和而不同”“兼收并蓄”“亲仁善邻”等中华优秀传统文化精髓,阐明了中国倡议深厚鲜明的文化底蕴,是中华文明在人工智能治理领域的最大彰显。

  提升全球人工智能治理能力

  倡议从全球人工智能治理的宏观与微观两个维度,为全面提升全球人工智能治理能力提供了切实可行的中国方案。其一,倡议在全球人工智能宏观治理的目标、理念、原则、能力等层面提出中国理念。在治理目标上,强调加快构建开放、公正、有效的全球人工智能治理机制,更好地促进人工智能技术造福人类,增进全人类共同福祉;在治理理念上,强调“以人为本”“智能向善”“负责任”等理念,从人工智能与人的关系、人工智能伦理等独特角度展现中国在人工智能治理方面的关键主张;在治理原则上,秉持安全与发展并重、相互尊重、平等互利、广泛参与、协商一致、公平性和非歧视性等基本原则,为协调各国人工智能治理冲突与全球人工智能治理变革提供公平、正确导向;在治理能力上,着力提升全球人工智能治理的权利保障能力、组织协调能力、国际协同能力、风险防范和化解能力以及可持续发展能力。其二,倡议在全球人工智能微观治理的对象、主体、工具、策略等层面提出中国方案。在治理对象上,倡议将人工智能数据采集、算法设计、模型训练、场景测试等技术开发,以及人工智能产品研发与应用等纳入重点治理视野,聚焦数据获取中个人隐私保护与数据安全、数据使用的真实性和准确性、算法的可解释性和可预测性等。在治理主体上,各国政府、国际组织、企业、科研院校、民间机构和公民个人等均是人工智能多元治理结构中的关键主体。在治理工具上,倡议强调建立健全人工智能领域法律和规章制度,完善标准规范与伦理体系,实现技术、法律、伦理、宣言、倡议共治,软硬法协同、内外法衔接。在治理策略上,以伦理先行、敏捷治理、分级分类的前瞻预见性、适应性治理机制,快速有效应对人工智能领域不断涌现的新风险点,实现“治理重点动态变化”与“治理政策动态调整”的融合。

  塑造全球人工智能治理新秩序

  倡议回应广大发展中国家诉求,推动全球人工智能治理体系朝着更加公正合理的方向发展。其一,倡议从全人类共同利益高度进行战略擘画,为加快构建智能时代的人类命运共同体贡献中国力量。倡议是数字文明时代中国为全人类提供的优质国际公共产品,强调在凝聚人工智能治理价值共识上应共同认可,在制定人工智能治理国际规则上应密切商议,在构建人工智能治理全球机制上应共同努力,在化解人工智能治理安全风险上应携手应对,形成更加公正普惠包容的全球人工智能治理秩序。倡议积极践行和平、发展、公平、正义、民主、自由的全人类共同价值,以“智能向善”谋求人类命运共同体的共同福祉,共赴数字文明新未来。其二,倡议回应广大发展中国家诉求,积极提升发展中国家参与全球人工智能治理的话语权。倡议是提升我国在全球人工智能领域话语权的重要举措,更是代表广大发展中国家在全球人工智能治理标准制定与规则表达上的普遍利益。倡议坚持发展中国家的角色定位,支持发展中国家更深入更广泛更有效地参与全球人工智能治理规则制定,充分反映发展中国家意见、利益和诉求,坚定不移致力于提高发展中国家在人工智能国际治理体系中的代表性和决策权,打破西方发达国家人工智能领域的话语霸权和规则垄断,发挥发展中国家在人工智能全球治理中的建设性作用。

  (作者系大连理工大学公共管理学院教授;大连理工大学马克思主义学院博士研究生)


The global significance of the Global AI Governance Initiative
2024-06-19 Author: Fei Yanying Zhao Kai Source: China Social Sciences Network-China Social Sciences Daily

General Secretary Xi Jinping emphasized that digital civilization should benefit people of all countries and promote the building of a community with a shared future for mankind. On October 18, 2023, President Xi Jinping delivered a keynote speech at the opening ceremony of the third Belt and Road Forum for International Cooperation, proposing the "Global AI Governance Initiative" (hereinafter referred to as the "Initiative"). The Initiative focuses on global cooperation in AI governance in the era of digital civilization, marking that my country has actively practiced the concept of a community with a shared future for mankind in the field of AI, and has taken concrete actions to implement the Global Development Initiative, the Global Security Initiative, and the Global Civilization Initiative. The Global AI Governance Initiative focuses on solving the historical question of "Where is human civilization going?" in the era of intelligence, contributing constructive ideas to global cooperation and effective governance of AI in the era of digital civilization, providing a blueprint for international discussions and rule-making in the field of AI, and sending out the "strong voice of China" leading global AI governance in the era of digital civilization.

Assisting the healthy development of global artificial intelligence

The initiative provides a conceptual guide for global artificial intelligence governance, and realizes global collaborative governance of artificial intelligence under the coordinated development and security of artificial intelligence. It not only points out the direction for the development of artificial intelligence in my country, but also injects a powerful Chinese force into ensuring the healthy, orderly and safe development of artificial intelligence in the world. First, the initiative highlights the concepts of "responsibility", "people-oriented" and "intelligent for good", providing value guidance for global artificial intelligence governance. Unlike some countries that strive to build an exclusive "small circle", the initiative adheres to the common, comprehensive, cooperative and sustainable security concept and the global governance concept of consultation, construction and sharing, advocates people-oriented, intelligent for good, diversified participation, open dialogue, co-construction and sharing, and win-win cooperation, and actively bridges the intelligence gap. The initiative provides constructive solutions for global artificial intelligence governance with a broader global vision, a wider range of artificial intelligence cooperation fields, a deeper level of artificial intelligence forward-looking layout, and a higher level of artificial intelligence governance capacity building. Second, coordinate the two key points of global artificial intelligence development and security. The initiative "adheres to the principle of giving equal importance to development and security", and takes agile governance and moderate constraints as principles, which not only overcomes the adverse effects of strong supervision on AI technology innovation, but also avoids the disorderly development of AI caused by weak supervision, and achieves a balance between AI technology development and security. Third, coordinate the relationship between national AI governance and global AI governance. The initiative fully reflects China's mature governance experience in legislation, strategic planning, ethical norms, and other aspects of AI in recent years, and also absorbs the outstanding achievements of multilateral parties in the international community, including the vast number of developing countries, in the field of AI governance. The initiative insists on coordinating the domestic and international situations, focusing on the country and the world, and advocates improving the global AI governance capabilities. It is necessary to build an open, healthy, and safe domestic AI responsible innovation ecosystem, and actively help other developing countries cross the intelligence gap and change the extremely unbalanced status quo in the field of global AI development.

Resolving global AI security risks

The initiative uses scientific thinking and methods, pays close attention to the tolerance and mutual learning between heterogeneous civilizations from the cultural confidence of Chinese civilization, gathers broad consensus for promoting the healthy development of global AI, and contributes Chinese wisdom to preventing and resolving global AI security risks. First, the initiative uses the "seven major thinking" to provide scientific methodological guidance for global AI governance. The initiative points out that we should achieve collaborative and systematic governance in the entire life cycle of AI data acquisition, algorithm design, technology development, product development and application; think about AI governance issues from a global perspective and raise them to a strategic level that concerns the fate of all mankind; have a deep insight into the historical position of the rapid development of AI technology and seize the historical opportunity of AI technology change; advocate giving equal importance to AI development and security, and taking both domestic and international considerations into account; innovate governance concepts, promote "intelligent good" with agile governance, and actively develop new AI governance technologies; explore the legalization of AI governance; emphasize that the development of AI should respect the sovereignty of other countries as the bottom line, and ensure that AI develops in a direction that is conducive to the progress of human civilization. Second, the initiative implements the "six insistences" and provides fundamental values ​​for global AI governance. Adhere to the people-first principle, and promote AI technology to assist, benefit and care for humanity with the concept of "people-oriented"; adhere to self-confidence and self-reliance, encourage all entities to work together to promote AI governance, and show a self-confident and self-reliant Chinese image; adhere to the principle of keeping the right path and making innovations, keep up with the development of cutting-edge AI technology, and deepen the understanding of the laws of technological development; adhere to problem-oriented, focus on preventing AI bias and discrimination, bridging the global intelligence gap and governance capacity gap, and other issues; adhere to the system concept, based on the global AI governance needs, and focusing on my country's AI governance practices, looking at the present and the future, seeking the unity of AI governance, and respecting the diversity of policies and practices of various countries; adhere to the world-minded, with both world and Chinese characteristics, oppose the artificial creation of AI development barriers, advocate dialogue and cooperation to promote AI for the benefit of mankind, and demonstrate a broad global vision and world-mindedness. Third, the proposal highlights the unique background of Chinese civilization's innovation, inclusiveness and peace, and endows global AI governance practices with Chinese wisdom from a civilizational perspective. The initiative adheres to the core propositions of "people-oriented", respecting the sovereignty of other countries, opposing ideological demarcation or building exclusive groups, and respecting the differences in AI policies and practices of various countries. It is imbued with the essence of China's excellent traditional culture such as "harmony but difference", "inclusiveness", and "being friendly to neighbors", which clarifies the profound and distinctive cultural heritage of the Chinese initiative and is the greatest manifestation of Chinese civilization in the field of AI governance.

Improving Global AI Governance Capabilities

The initiative provides a practical Chinese solution for comprehensively improving global AI governance capabilities from the macro and micro dimensions of global AI governance. First, the initiative proposes Chinese concepts at the levels of goals, concepts, principles, and capabilities of global AI macro governance. In terms of governance goals, it emphasizes accelerating the construction of an open, fair and effective global AI governance mechanism to better promote AI technology to benefit mankind and enhance the common well-being of all mankind; in terms of governance concepts, it emphasizes concepts such as "people-oriented", "intelligent for good" and "responsible", and presents China's key propositions in AI governance from unique perspectives such as the relationship between AI and people and AI ethics; in terms of governance principles, it adheres to the basic principles of balancing security and development, mutual respect, equality and mutual benefit, broad participation, consensus, fairness and non-discrimination, and provides a fair and correct direction for coordinating AI governance conflicts among countries and global AI governance reforms; in terms of governance capabilities, it focuses on improving the rights protection capabilities, organizational coordination capabilities, international coordination capabilities, risk prevention and mitigation capabilities, and sustainable development capabilities of global AI governance. Second, it advocates proposing Chinese solutions at the levels of objects, subjects, tools, and strategies for global AI micro-governance. In terms of governance objects, it advocates incorporating AI data collection, algorithm design, model training, scenario testing and other technical developments, as well as AI product development and application, into the key governance vision, focusing on personal privacy protection and data security in data acquisition, the authenticity and accuracy of data use, and the interpretability and predictability of algorithms. In terms of governance subjects, governments, international organizations, enterprises, scientific research institutions, non-governmental organizations and individual citizens are all key subjects in the multi-governance structure of artificial intelligence. In terms of governance tools, the initiative emphasizes the establishment and improvement of laws and regulations in the field of artificial intelligence, the improvement of standards and ethical systems, the realization of co-governance of technology, law, ethics, declarations, and initiatives, the coordination of soft and hard laws, and the connection of internal and external laws. In terms of governance strategy, with ethics first, agile governance, and a forward-looking and adaptive governance mechanism with hierarchical classification, we can quickly and effectively respond to the emerging new risks in the field of artificial intelligence and realize the integration of "dynamic changes in governance priorities" and "dynamic adjustments in governance policies".

Shaping a new order for global artificial intelligence governance

The initiative responds to the demands of developing countries and promotes the development of the global artificial intelligence governance system in a more fair and reasonable direction. First, the initiative makes strategic plans from the height of the common interests of all mankind and contributes China's strength to accelerate the construction of a community with a shared future for mankind in the intelligent era. The Initiative is a high-quality international public product provided by China for all mankind in the era of digital civilization. It emphasizes that there should be common recognition in building consensus on the value of AI governance, close consultation in formulating international rules for AI governance, joint efforts in building a global mechanism for AI governance, and joint response in resolving security risks in AI governance, so as to form a more fair, inclusive and inclusive global AI governance order. The Initiative actively practices the common values ​​of peace, development, fairness, justice, democracy and freedom for all mankind, seeks the common welfare of the community with a shared future for mankind with "intelligent goodness", and jointly moves towards a new future of digital civilization. Second, the Initiative responds to the demands of the vast majority of developing countries and actively enhances the voice of developing countries in participating in global AI governance. The Initiative is an important measure to enhance my country's voice in the field of global AI, and it also represents the general interests of the vast majority of developing countries in the formulation of global AI governance standards and the expression of rules. The Initiative adheres to the role of developing countries, supports developing countries to participate more deeply, widely and effectively in the formulation of global AI governance rules, fully reflects the opinions, interests and demands of developing countries, and is firmly committed to improving the representation and decision-making power of developing countries in the international AI governance system, breaking the discourse hegemony and rule monopoly of Western developed countries in the field of AI, and playing a constructive role of developing countries in the global governance of AI.

(The author is a professor at the School of Public Administration of Dalian University of Technology; a doctoral student at the School of Marxism of Dalian University of Technology)

No comments:

Post a Comment