Open Forum #33 Building an International AI Cooperation Ecosystem

Summary

This open forum, hosted by the Bureau of International Cooperation of the Cyberspace Administration of China, focused on building an international AI cooperation ecosystem based on extensive consultation, joint contribution, and shared benefits. The discussion brought together experts from various countries and organizations to address AI governance and development challenges. Director-General Qi Xiaoxia opened by highlighting China’s leadership in AI innovation, noting that over 60% of global AI patents come from China and more than 430 generative AI service models are registered there. She proposed three key areas for international cooperation: ensuring sustainable development and technology for good, making proactive responses to AI risks while ensuring safety and controllability, and fostering consensus on AI governance through collaborative approaches.

Several speakers emphasized that AI has evolved from a mysterious technology to a commodity accessible to many developers, with Jovan Kurbalija citing the “DeepSeek moment” as evidence that effective AI can be developed with limited resources through smart algorithms rather than just processing power. Wolfgang Klauweiter discussed the explosion of AI governance frameworks globally, from OECD principles to the EU AI Act, while advocating for the United Nations to take a leading role in AI governance since it’s the only forum that includes all countries. Multiple speakers stressed the importance of multi-stakeholder approaches, drawing parallels to successful Internet governance models.

The discussion highlighted various national and regional experiences, including Mexico’s rights-based regulatory framework, Shanghai’s comprehensive AI industrial ecosystem development, and China Mobile’s AI Plus Action Plan. Participants consistently emphasized the need for inclusive, ethical AI development that bridges digital divides and serves humanity’s greatest aspirations rather than deepest fears. The forum concluded with a consensus that international cooperation and shared governance frameworks are essential for managing AI’s transformative impact on society.

Keypoints

## Major Discussion Points:

– **AI as a Transformative Global Force**: Multiple speakers emphasized that AI has evolved from a mysterious technology to a commodity that is reshaping economies, industries, and governance systems worldwide. The “DeepSeek moment” was highlighted as a turning point showing that effective AI can be developed with limited resources and smart algorithms rather than just massive computing power.

– **Need for International Cooperation and UN Leadership**: There was strong consensus that AI governance requires global collaboration rather than isolated national approaches. Speakers advocated for the United Nations to take a leading role in AI governance, building on existing internet governance frameworks and bringing all countries to the table through inclusive multilateral processes.

– **Balancing Innovation with Risk Management**: The discussion addressed the fundamental challenge of fostering AI innovation while managing risks such as data privacy, algorithmic bias, digital divides, and potential harm to human dignity. Speakers emphasized the need for regulatory frameworks that don’t stifle innovation but ensure safety, controllability, and fairness.

– **Multi-stakeholder Approach and Inclusive Development**: Drawing parallels to internet governance, speakers stressed that no single stakeholder has all the answers. They advocated for bringing together governments, academia, private sector, civil society, and international organizations to ensure AI development serves all of humanity, particularly addressing the needs of developing countries and marginalized communities.

– **Practical Implementation and Regional Experiences**: The forum showcased concrete examples of AI governance and development, including China’s regulatory measures, Shanghai’s innovation ecosystem, Mexico’s rights-based approach, and various international initiatives. Speakers shared specific policies, frameworks, and collaborative mechanisms being implemented in their respective regions.

## Overall Purpose:

The discussion aimed to explore how to build an international AI cooperation ecosystem based on “extensive consultation, joint contribution, and shared benefits.” The forum sought to share experiences, best practices, and policy approaches for AI governance while fostering dialogue between different stakeholders and regions to develop collaborative frameworks for responsible AI development.

## Overall Tone:

The discussion maintained a consistently collaborative and optimistic tone throughout. Speakers were respectful and constructive, focusing on shared challenges and opportunities rather than competitive or adversarial positions. There was a notable emphasis on inclusivity, with frequent references to ensuring AI benefits all of humanity. The tone remained professional and forward-looking, with speakers building upon each other’s points and offering concrete proposals for international cooperation. The moderator maintained good control of the session, keeping discussions focused and time-efficient.

Speakers

**Speakers from the provided list:**

– **Chin Yik Chan**: Moderator, Professor Xu Peiqi (also goes by Patrick), moderating the forum hosted by the Bureau of International Cooperation of the Cyberspace Administration of China

– **Qi Xiaoxia**: Director-General from the Cyberspace Administration of China, representing the host of the forum, Bureau of International Cooperation

– **Jovan Kurbalija**: From the Diplo Foundation

– **Wolfgang Klauweiter**: Professor from the University of Akros

– **Dai Wei**: From the Internet Society of China

– **Sajid Rahman**: Board member of ICANN

– **Ricardo Pelayo**: Professor from the Banking and Commercial School of Mexico

– **Wushu Yan**: From China Mobile Research Institute

– **Participant**: Delivered a video message (identity not clearly specified in transcript, but mentioned as delivering remarks about AI innovation and cooperation ecosystem)

**Additional speakers:**

– **Professor Dai Li Na**: From the Shanghai Academy of Social Sciences (mentioned in the agenda but appears to have spoken as “Participant” in one section)

– **Luiky Gambadela**: From China-EU (non-profit international association), delivered a short video message (mentioned in agenda but may correspond to one of the video messages in transcript)

# International AI Cooperation Ecosystem Forum: Discussion Report

## Executive Summary

This open forum, hosted by the Bureau of International Cooperation of the Cyberspace Administration of China, brought together international experts to explore building an AI cooperation ecosystem. The discussion was structured in two sessions: fostering a favorable policy environment for AI development, and building AI innovation and cooperation ecosystems. Representatives from China, Mexico, Europe, and international organizations addressed AI governance challenges and cooperation opportunities.

Key themes included the democratization of AI technology, the need for multi-stakeholder governance approaches, building upon existing internet governance frameworks, and ensuring inclusive development that addresses digital divides and human rights concerns.

## Opening Remarks

**Director-General Qi Xiaoxia** from the Cyberspace Administration of China opened the forum by highlighting China’s position in global AI development, noting that over 60% of the world’s AI patents originate from China, with more than 430 generative AI service models registered domestically and over 2,800 million registered users.

She outlined three areas for international cooperation: ensuring sustainable development and technology for good, making proactive responses to AI risks while ensuring safety and controllability, and fostering consensus on AI governance. Qi emphasized China’s support for the Global AI Governance Initiative and noted China’s interim measures for generative AI services management.

## Session 1: Fostering Favorable Policy Environment

### The Democratization of AI Technology

**Jovan Kurbalija** from the Diplo Foundation highlighted a fundamental shift in AI accessibility, referencing the “DeepSeek moment” on January 20th when it was demonstrated that effective AI could be developed with limited resources through smart algorithms rather than massive processing power. He argued that AI has transformed from “mysterious technology in the hands of a few developers and top labs with huge investment of money” to a commodity accessible to everyone.

Kurbalija emphasized the importance of preserving community knowledge and avoiding dependence on large platforms. He suggested developing standardization around AI weights to enable international cooperation and proposed that the UN develop open source AI models with countries contributing their national AI models.

### Global Governance Frameworks

**Professor Wolfgang Klauweiter** from the University of Akros analyzed the proliferation of AI governance frameworks globally, including OECD principles, the EU AI Act, UNESCO recommendations, and various national strategies. He advocated for the United Nations to take a leading role in AI governance as “the only forum that includes all countries.”

Klauweiter noted that even the complex European AI Act faces implementation challenges and suggested that AI governance could build upon existing internet governance frameworks, observing that “AI is just using the Internet” and therefore existing multi-stakeholder models could be adapted.

### Multi-Stakeholder Approaches

**Sajid Rahman**, ICANN board member, emphasized that AI’s growth is “unprecedented compared to previous technological waves.” He stressed the urgency of cooperation, stating: “the question is not whether we should cooperate. It is whether we do fast enough, broadly enough, and wisely enough to steer AI towards humanity’s greatest aspiration, rather than deepest fears.”

Rahman advocated for multi-stakeholder frameworks drawing from internet governance experience, emphasizing the need to address digital divide issues and ensure equal access to prevent AI from deepening existing inequalities.

### Regional Perspectives

**Professor Ricardo Pelayo** from the Banking and Commercial School of Mexico discussed Latin American approaches, noting that while Mexico lacks specific AI legislation, it has existing frameworks for data protection and children’s rights. He emphasized that “AI development must prioritize human rights, equity, and not come at expense of human dignity.”

Pelayo announced plans to present comprehensive public policy proposals to the Mexican Congress and stressed the importance of incorporating local languages, cultures, and community aspirations into AI development.

## Session 2: Building AI Innovation and Cooperation Ecosystem

### Chinese Implementation Models

**Dai Wei** from the Internet Society of China highlighted the organization’s cooperation with over 20 global organizations, including the Internet Governance Forum, ICANN, and UNESCO.

**Professor Dai Li Na** from the Shanghai Academy of Social Sciences presented a comprehensive case study of Shanghai’s AI industrial ecosystem development, outlining five key pathways: government guidance, industrial chain improvement, academia-research integration, industrial integration, and international cooperation. She emphasized the importance of public-private partnerships in accelerating development and translating ideas into real-world impact.

**Wu Shuyan** from China Mobile Research Institute presented the company’s AI Plus Action Plan, highlighting their JiuTian Foundation model and the AI Plus Pro Global Acquisition Alliance for international AI adoption. The presentation demonstrated how major telecommunications companies are positioning themselves as AI providers and operators with global solutions.

### Video Message Contribution

A video message participant discussed international cooperation mechanisms and the importance of inclusive approaches to AI governance, though the specific identity of this speaker was not clearly established in the discussion.

## Key Areas of Agreement

Participants demonstrated consensus on several fundamental principles:

– The necessity of multi-stakeholder approaches to AI governance

– Support for UN leadership in global AI governance

– The importance of international cooperation in addressing AI challenges

– Building upon existing internet governance frameworks rather than creating entirely new structures

– Balancing AI development with human rights protection and digital divide concerns

## Concrete Proposals and Action Items

Several specific proposals emerged from the discussion:

– Developing standardization around AI weights for international cooperation

– UN development of open source AI models with national contributions

– Using Sustainable Development Goals to guide AI development and governance

– Strengthening ties between global industry organizations to share best practices

– Presenting comprehensive AI policy proposals to national legislatures

## Implementation Challenges

The forum identified several ongoing challenges:

– Implementing complex risk-based regulatory systems without stifling innovation

– Addressing regulatory gaps in countries lacking specific AI legislation

– Ensuring transparency of AI algorithms from ethical and regulatory perspectives

– Balancing technological development with national security interests

– Addressing global challenges like data privacy, algorithmic bias, and digital divides

## Conclusion

The forum demonstrated broad international consensus on fundamental AI governance principles while highlighting the need for continued dialogue on implementation approaches. The emphasis on multi-stakeholder cooperation, building upon existing frameworks, and ensuring inclusive development provides a foundation for advancing international AI governance. The discussion reflected both urgency about AI’s transformative impact and pragmatism about adapting existing governance structures to meet new challenges.

The combination of policy frameworks, industry initiatives, and international cooperation mechanisms discussed suggests multiple pathways for advancing responsible AI development that serves global interests while respecting national priorities and local community needs.

Chin Yik Chan: Okay. Can you hear me? Yes, yes. Okay. Good afternoon, friends and colleagues, and welcome to this open forum. It is hosted by the Bureau of International Cooperation of the Cyberspace Administration of China, or the CEC. I am the moderator, Professor Xu Peiqi. You can also call me Patrick, my nickname. And my suggestion is that you use that device on your side. We are with Channel 5. So Channel 5, we are with. The topic of this open forum is related to artificial intelligence governance. Our specific topic is building an international AI cooperation ecosystem based on extensive consultation, joint contribution, and shared benefits. And we have 60 minutes. Each speaker or panelist has five minutes. Our speakers, according to the sequence, are from China. In the agenda will be firstly the Director-General Qi Xiaoxia from the Cyberspace Administration of China. She is representing the host of this forum. She will be followed by Yuvan Kapelidze from the Diplo Foundation, and Professor Wolfgang Klauerwachter from the University of Akros, and Mr. Luiky Gambadela from China-EU, which is a non-profit international association, and he is not on site. He will deliver a short video message. And Mr. Dai Wei from the Internet Society of China. And then we have Mr. Sajid Rahman from ICANN, and also Professor Ricardo Pelayo from the Banking and Commercial School of Mexico, and Professor Dai Li Na from the Shanghai Academy of Social Sciences. And finally, we have Ms. Wu Shuyan from China Mobile Research Institute. Now let’s welcome Ms. Qi Xiaoxia to give opening remarks.

Qi Xiaoxia: Thank you, Professor, distinguished guests, ladies and gentlemen, friends, good afternoon. I’m delighted to meet all of you in Norway and join the discussion on AI governance, development, and international cooperation. On behalf of the Bureau of International Cooperation, Cyberspace Administration of China, I wish to extend a warm welcome and heartfelt appreciation to all guests, both present and online today. Nowadays, with the continued technological breakthroughs, emergence of new business models, and accelerated expansion of applications, artificial intelligence has become a strategic technology spearheading a new round of scientific and technological revolution and industrial transformation. It is a critical force driving global technological and industrial development, exerting a profound impact on human production and life. At the same time, it presents unprecedented challenges to humanity. AI governance has become a common concern of countries worldwide. China has always placed great emphasis on AI development and governance, underscoring the necessity of jointly promoting development and application of AI technologies while ensuring safety, reliability, controllability and fairness throughout its development process so as to fully harness AI’s potential to bring greater benefits to humanity. I’d like to take this opportunity to introduce China’s practice and experience in this regard. Firstly, promoting innovation and development of AI industry. China is at the forefront of AI technology innovation in the world, with more than 60% of the world’s granted AI patents coming from China. As of now, more than 430 generative AI service models have been registered and put online in China, with over 2,800 million registered users. China has become an important market for AI technology innovation and industry application. Secondly, strengthening AI safety governance. China has released the interim measures for the management of generative AI services, marking the first legislative effort of its kind globally, and made public the AI safety governance framework providing better safeguards for the healthy development of AI. Thirdly, deepening international cooperation on AI governance. China put forth and implemented the Global AI Governance Initiative, supports fair competition between domestic and foreign AI companies, and stays committed. to fostering an ecosystem conducive to the healthy and sustainable development of large-language model. Tackling new challenges brought by AI requires joint efforts of governments, international organizations, enterprises, and scientific institutions to develop open, fair, and efficient AI governing mechanisms so as to minimize possible negative impacts while fully amplifying and accelerating the positive effects of AI. We look forward to working with other parties to intensify exchanges and cooperation in AI development and governance. We are committed to governing AI for all, creating an open and inclusive environment for the development of AI so as to make emerging technologies represented by AI to serve the greater good of humanity. With this in mind, I wish to propose efforts in three areas. First, staying focused on sustainable development and ensuring technology for good. It is important to help developing countries enhance capacity building, increase their representation and voice in global AI governance, ensure equal rights, equal opportunities, and equal rules for all countries in AI development and governance. International cooperation with the assistance to developing countries should be carried out to bridge the intelligence divide and narrow the gap in governance capacity. Second, making proactive responses to risks and challenges and ensuring AI safety and controllability. Safety and controllability is the prerequisite for the healthy and orderly development of AI and the precondition for AI technologies to benefit, not harm, human beings. China stands ready. to work with all countries to prevent and manage risks, ensure AI technologies are safe, reliable, controllable, and fair, and develop AI in a way that is beneficial to the progress of human civilizations. Third, fostering consensus on AI governance and deepening collaborative governance. IGF is the important platform under the United Nations. The leading role of the UN in international AI governance should be supported with full respect for the differences among countries at the policy and practical levels. It is important to encourage the active participation of multi-stakeholders, build consensus on international AI governance, and ensure all countries can share intelligence dividends. Thank you for your attention. I wish the forum a full success.

Chin Yik Chan: Thank you very much, Ms. Director General. You have introduced the Chinese experience and the practice of AI governance and development in three areas, and also you have made three proposals. So now we move on to session one. The topic of session one is about fostering a favorable policy environment for AI development, and we’re going to start from you, Ivan.

Jovan Kurbalija: Thank you. Thank you, Patrick, Director General, dear colleagues and friends. It’s my honor to be a participant at this panel, and I would like to start from one statistic. I didn’t remember completely the number which the Director General mentioned, but it was something like 450 generative AI. 430. Yes. There is a saying that in China, a new model is developed every day, and this is the first point that I would like to forward here. AI is becoming a commodity. Thank you. Only three years ago, AI was a mysterious technology in the hands of a few developers and the top labs with huge investment of money. That has changed. And most of our governance thinking is related to what happened two years ago, three years ago, including the famous letter that AI will destroy humanity and other elements. It’s a time to revisit and to see what is basically our governance should regulate and what should be the governance function that we should cover by governance models. This is the first point. AI is commodity. You can create an agent in five minutes and that’s basically given. Here another point is that I call it deep seek moment, 20th of January, where deep seek was released for the fraction of the cost. Although there are controversies how costly it was, but let’s say it was much, much less than major platform. It was shown that you can develop AI from the foundational model with the limited funding with limited NVIDIA cards. And it was the critical point because till that moment, the idea was the more NVIDIA cards or whatever processing power you have, the better AI will create. For the first time, somebody said, no, you don’t need just the sheer processing power. You need a smart algorithms, smart solution and smart architecture. Therefore, this has changed completely AI governance landscape. And I will return to director general’s few points, concrete points, how we can revisit that core notions and create, I would say revitalize or new approaches to AI governance nationally and globally. First one is about SDGs, SDGs should be used, AI should be used for achieving SDGs, but we should also use SDGs to govern AI. SDGs are the latest. and the most comprehensive codification of the core human values and priorities agreed by all UN member states. Why not to create AI systems that will implement each of SDGs? Now with automating coding like cursor and the github and other platforms it’s doable. That’s the first concrete proposition. Second concrete proposition is that we need bottom-up AI. This empowerment of 430 platforms and platforms all over the world in the United States, in Europe. We are empowering us to have fast AI models on our mobiles and it provides unique chance for internet governance community and other communities that we can preserve our knowledge. We can preserve our knowledge and not pass to big platforms and text giants wherever they are located. I’m not now referring to any geopolitical context but wherever they are located. Knowledge is what defines us as a human beings. This is definition of our human dignity, our personal knowledge and knowledge of our communities. Interestingly enough in WSIS documents you have many references to knowledge society and knowledge. Somehow knowledge is cleaned from the political discourse of the last 20 years. You don’t have any more knowledge, you have only data. And what we are speaking about AI is the knowledge. This is just one idea for young researchers to see how we clean the knowledge from the discourse. This is the first point Director General to have SDGs and to have a revitalization of the knowledge and bottom-up AI. And I will conclude with two more concrete proposals. One is to insist on the share weights and standards for sharing the weights. This is weights in AI are basically the way of creating AI. This is how international cooperation will be conducted. Can we use the weights from let’s say DeepSeek, to be shared with Mistral and other platforms. For the time being, not. We already went a step in this trap with social media, where when you are on one social media platform, you cannot bring your contacts on the other social media platform. Let us learn something from our mistakes. When I say mistakes, I say humanity as a whole. We should have the standardization around the AI ways. And last very concrete proposal, and with that I will conclude, is to insist that the UN develop open source AI models for the UN activities. Countries can contribute AI models. When you go to New York, you can see sculptures, paintings as national contributions. For the new AI, for the new UN, especially at this critical moment, where, as you know, the UN is in crisis, we need to rethink what we contribute as countries. And that would be great if China, the United States, other countries, small countries, contribute their AI models for the future UN. Those are just a few concrete proposals inspired by your opening remark. Thank you.

Chin Yik Chan: Thank you very much, Jovan. So you firstly argue that AI is a commodity, and secondly you talk about the characteristics of a deep-seated moment, and finally you made several proposals. Shall we move on to Professor Wolfgang Klauweiter?

Wolfgang Klauweiter: Yeah, thank you very much. Thank you also for the invitation. And as Jovan just said, you know, AI is not any more a mysterious science fiction. It’s part of our daily life. And like everything in our daily life, we have to have some guardrails, we have to have some orientations, you know. Our daily life is, quote-unquote, regulated by legal norms, by just interpersonal behavior norms, by moral norms, and other norms. So it’s not a surprise that we, if we discuss AI governance, we speak about norms, principles, rules, you know, how to organize this. And in the last five, six years, we have seen an explosion of discussions around these norms, principles, and things like that. And it’s difficult to get the overview. It started already nearly 10 years ago in the OECD, which drafted the first set of principles in 2009. which fortunately were adopted in by the G20, where China is a member, and from this then we did see a number of regulatory initiatives, in particular in Europe, where the European Union adopted a first set of legal binding norms, the Council of Europe adopted a convention on the protection of human rights and AI, and we did see UNESCO, which adopted a recommendation, which is not legally binding, for the ethical dimension of AI, and we did see a number of mixed processes, like the Pledge Lake process, which was started by the United Kingdom two or three years ago. We see the initiative of the so-called Hiroshima process of the G7 countries, and finally we did see the discussion in the United Nations in the last two or three years with the high-level panel, and now the proposals which are included in the Global Digital Compact. You know, all this are partly complementary initiatives, but the big question behind this is, you know, what do you really want to regulate? What is AI governance? So that means everybody agrees we need something, but there is partly a confusion because you have a dilemma, while on the one hand you want to stimulate innovation, you want to have new developments, so, and the risk is if you regulate it, then you can probably strangulate innovation, you can stop this, and to find the right balance is rather difficult, you know. From the European perspective, this was the problem with the European Union AI Act, where they introduced four categories, so the so-called risk-based approach, and they said, okay, for applications which violate human dignity, this should be forbidden. For risky applications, we have to have very specific binding rules. For activities which are low risk, we are much more flexible, and for applications with no risk, so everybody goes, so that means we have no regulation. The problem is if you introduce such a complex system, how to implement it, and in the European now the discussion has started about the simplification of the regulation which was just adopted. So that’s really a challenge and nobody has a real answer. But that’s not a problem for Europe alone, that’s not a problem for China alone, and it’s not a problem for the United States alone, as we heard this morning in the opening, that the new administration wants to remove all kinds of regulation just to let the free market decide. This is a global problem and I can only support what Madame Xi has said in the opening. If it’s a global problem, we need all countries on the table, and only the United Nations offers this opportunity. So that means the two initiatives which are now on the table in the United Nations, the Scientific Panel and the AI Governance Dialogue, this could take the lead because in all the other initiatives you have a lot of exclusion, many countries are not involved. And as German has said, this is a problem for all countries, small and big countries. And so far one message here from the IGF and from this workshop could be that we have to strengthen the leading role of the UN in AI governance. But let me add here a final point. There is no need to reinvent the wheel because we are on the Internet Governance Forum and I was involved in discussions in the last couple of days about what is the difference between Internet governance and AI governance. Is this totally new or not? To a certain degree, AI governance or AI is just using the Internet. So that means you can build on what has been developed in the last 20 years in the Internet Governance Forum with the Internet and you can use the definition for Internet governance also for AI governance. So that means to build on what you have. And the third element is you need a holistic approach. And this is relevant also for AI governance. So that means no reinvention of the wheel, use the United Nations as a leading institution and look into the future, do not strangulate IE development. Thank you.

Chin Yik Chan: ≫ Thank you very much, Wolfgang. And I would like to ask you a few questions. First of all, I would like to ask you a few questions about AI governance. And I would like to ask you a few questions about AI development. Thank you. ≫ Thank you very much, Wolfgang. So you observe the emergence of international regimes and frameworks to regulate AI. Our third speaker will deliver a video message and his name is Giorgio. Giorgio, can you please tell us about AI and how it is implemented in the world?

Participant: ≫ Distinguished guests, dear friends, it is a great honor to speak to you today on a topic that is reshaping not only our economies, but the very essence of human progress, artificial intelligence. And I would like to start with the first point which is the need to build a robust AI innovation and cooperation ecosystem has never been more urgent and at the same time promising. AI is not just a tool. It is a transformative force, reshaping every aspect of our lives. AI is set to play such a critical and 1992 major role in public life, education, transportation, agriculture and climate action. be unlocked if we create a framework that fosters innovation while promoting cooperation across borders, sectors and disciplines. First, we must recognize that innovation does not happen in isolation. Breakthroughs arise from diversity and connection of ideas, talents, cultures and technologies. To harness this potential, we need an ecosystem that actively cultivates these diverse elements and enables them to interact seamlessly. Creating a thriving ecosystem as such requires substantial investment in education, research and digital infrastructure. Equally important is supporting startups, nurturing diverse talent pools and strengthening the links between research, institutions and industry. Public-private partnerships are not optional, but an imperative to accelerate development and translate ideas into real-world impact. Second, cooperation must go beyond national agendas. AI is global by nature. Data flows, algorithms and applications do not respect borders. That is why we must build bridges, not walls. International cooperation on ethical standards, regulatory framework and data sharing mechanism is vital. We must avoid the trap of technological protectionism and instead embrace open innovation with trust and transparency at its core. Europe, for instance, has taken important steps with its AI Act, placing human-centered values at the heart of its vision. Yet this is only the beginning. Initiatives ranging from leading academic hubs like Stanford Human-Centered AI to national strategies such as China AI Plan and multinational programs like the EU Invest AI and the AI Continent Action Plan. Each play unique roles in advancing AI, but these efforts should not operate in silos. We need dialogue between ecosystems, shared forums, joint projects and trust building mechanisms that foster mutual understanding and interoperability. Third, our ecosystem has a responsibility and a tremendous opportunity to support strategic sectors where AI can deliver rapid, scalable and inclusive benefits such as precision medicine, renewable energy optimization, and digital agriculture. This domain demands international coordination, from research to standards to deployment. Only through cooperation can we tackle global challenges like pandemics, food insecurity or climate resilience. Fourth, for innovation to truly serve the world, it must be inclusive. We must avoid an AI future that deepens inequality or concentrates power in a few regions or companies. That means giving space and support to a diverse range of actors, including SMEs, startups and researchers, especially in the Global South. It means promoting open source tools, multilingual models and accessible platforms that allow more actors to benefit and contribute, empowering a global community to shape and share the benefits of AI. Ladies and gentlemen, for the first time in human history, machines are learning. What we teach them, how we govern them, will define the future of our civilization. This responsibility requires more than a quick fix. It demands building an AI, innovation and cooperative ecosystem. Such as Endeavor is not a short-term project. It is a generational mission. It calls for boldness, imagination and unwavering commitment. Let us build an AI future where knowledge is shared, ethics are global and progress benefits all humanity. Thank you.

Chin Yik Chan: That’s all about the framework of cooperation and innovation in AI. Now let’s move on site. Our next speaker will be Mr. Dai Wei, who is sitting beside me. He is from the Internet Society of China.

Dai Wei: Distinguished guests, ladies and gentlemen, good day to you all. I’m delighted to join you in this United Nations Internet Governance Forum. On behalf of the Internet Society of China, I would like to extend my congratulations to the workshop. Just now Mr. Jhova talked about the DeepSeek movement. Mr. Wolfgang talked about AI governance and knowledge. Mr. Rijie talked about the impact of AI cooperation. Today I will talk about building an international AI cooperation ecosystem based on extensive consultation, joint contribution and shared benefits. In recent years, artificial intelligence has become a driving force for technological revolution and industrial transformation. It is reshaping economies, industries and governance systems. At the same time, as AI becomes more integrated into our society’s risks, we are also seeing more challenges from data security risks to algorithmic bias, from social inequality to technological monopolies. And we did a lot of things for the… people with disabilities and the elderly to narrow the digital divide. So that’s why building an international AI cooperation ecosystem is no longer just an option. China has long supported a more inclusive global governance. As a key player in China’s Internet civil society, Internet Society of China remains committed to openness, collaboration, and engagement with partners around the world in the field of AI. We have cooperation with over 20 organizations and institutions worldwide, such as including IGF, ICANN, UNESCO, and KISA is a career organization for private information protection. So we believe that only by uniting all stakeholders, governments, industry, universities, and civil society can we harness the potential of AI while managing its risks. As part of our efforts to promote responsible Internet and AI development, we also serve as the secretary of the China Internet Governance Forum. Through this platform, we bring together voices from across sectors, both domestic and international, to share ideas and tackle challenges and collaborate on building a more inclusive digital future. We have set up an AI working committee that brings together experts from universities, research institutions, and businesses. We host salons, forums, and competitions to foster innovation and knowledge sharing. At the same time, we are developing technical standards and talent training systems to support the long-term healthy growth of the industry. So this afternoon, I heard some workshops that discuss the role of international cooperation. So the role of international cooperation is very important. So looking to the future, Internet Society of China will be committed to the development and governance of AI. We aim to strengthen ties with global industry organizations, share experiences, promote best practices, and respond to the complex challenges. Our goal is to help build a global AI ecosystem that is open, inclusive, and innovative. and mutually beneficial. We believe that through our shared efforts, AI can become a force for good, advancing global economic and social development, and creating a better future for all. Thank you.

Chin Yik Chan: Thank you very much, Mr. Dawei. So you talked about how, for example, the Internet Society of China is very widely connected to the global community and the global mighty stakeholders, and then you also mentioned about the AI Working Committee of the Internet Society of China. That is a typical feature of the institutional arrangement in China of establishing AI Working Committees. And so we can now move on to the second session. If time permits, we will come back for discussion among the panelists. Our second session topic is about building AI innovation and cooperation ecosystem, and we are going to start from Mr. Sajid Rahman. He is a board member of ICANN. Sajid.

Sajid Rahman: Thank you, and good afternoon. You know, it’s a great pleasure to speak about something which is not only shaping our lives, but also demanding combined stewardship from different stakeholders of the society. When Internet started, the people involved had the farsightedness to realize that no one stakeholder has all the answers, and so multi-stakeholderism evolved, and which included all the opinions and the inputs and the combined decision-making that helped Internet to grow where it is today. The thing about AI is that, you know, and I have a few friends who are in the investment world and who have invested in a lot of companies into what we call Internet 2.0, the social medias and etc., and now they’re investing in AI companies, and they keep telling me that the growth of AI and the way it is transforming every day is something they have not seen in their previous investment life. So, the acceleration is dramatic, and it is going to have a far-reaching impact than what we have experienced before. Now, when we are talking about developing an AI innovation and cooperation ecosystem, the good news is that there is actually a model, and the model is, you know, this whole multi-stakeholder framework that we talk about, which brings the technical expertise, the public interest, and global perspective to foster cooperation, trust, and accountability in how AI is deployed and developed. The important thing here is that, you know, it is not only the investors, it’s not only the regulators, it is the whole part of different parts of the society. So, let me give an example. We talked about, you know, the dipstick moment came up a few times. We talked about Cursor. We see how open AI is rolling out different models almost every day. two, three months. We’re seeing how Anthropic and Perplexity and Databricks and, you know, NVIDIA, you know, all the different players in this field are coming together. But at the same time, it is very important to understand the significant impact of academia and open research communities who provide the scientific foundations. In fact, what we see as the manifestation of AI and what we think has suddenly happened actually didn’t happen. There had been a deep scientific research in academia which has been used to build what we see today. So we need to bring all the academia and open research, you know, communities together. We need to obviously bring in the public and private sector actors, especially the private sectors who build technologies, build companies, and ensure that these are monetized properly. We obviously need to bring in the government and international bodies who set the regulatory and ethical framework to guide responsible deployment. And it is very important at this point because the transparency of AI algorithms is becoming a question and a challenge, whether from an ethical perspective or from how it is getting treated by different regulatory bodies. So that remains, you know, a deep point of discussion as we talk about innovation and cooperation. And then, of course, the civil society. And we keep talking about digital divide. We saw the digital divide in Internet 2.0. The question is, how deep will the digital divide be with AI coming in, right? So how do we bring all the civil society and the marginalized communities also part of the cooperation and ecosystem? Now, one of the things which I think is very good, and it seems, you know, while the governance of AI is not final yet, the ink is not dry, it is evolving, but we are seeing a significant shift in the right directions. You know, initiatives including UN Secretary General’s high-level advisory body on AI, or the OECD’s AI principles, the Global Partnership on AI, Councils of Europe’s work on AI Convention. These are all sort of incorporating the multi-stakeholder process as a movement towards the governance of this, the governance of AI. From an ICANN hat, you know, as you know that we are very much a firm believer of multi-stakeholderism, and this is something, you know, while AI governance is beyond our limit, but at the end of the day, AI is built on, as Wolfgang touched upon, AI is built on Internet. And one of the things that we do is ensuring a stable, secure DNS, and the digital world, and AI, so that the AI can scale safely on the platform of a scalable Internet. One of the things I want to close is, you know, the question is not whether we should cooperate. It is whether we do fast enough, broadly enough, and wisely enough. To steer AI towards humanity’s greatest aspiration, rather than deepest fears. You know, there are, it’s like any other big technology that suddenly hits the civilization. And the question is, do we grab that technology and really steer it for the development and the improvement of every human being on Earth, or do we create more silos? We obviously don’t want to go to the next one. We don’t want it to be a fear, but we want it to be our greatest aspirations. So, to build an AI ecosystem that is innovative, inclusive, and sustainable, we need courage, we need humility, and we need commitment. Let this be the decade that we get this right. We won’t have a second chance. Thank you.

Chin Yik Chan: Thank you very much, Sajid. So, you emphasize that every stakeholder should be brought together in terms of global AI governance, which is similar to Internet governance. So, then we are going to have Professor Ricardo Pelayo.

Ricardo Pelayo: Hi, good afternoon. It’s an honor to share with you this reflection on building an ecosystem of innovation and cooperation in artificial intelligence from the legal and public policy context of Mexico. We are living in a crucial moment. As some colleagues said before, artificial intelligence is no longer a promise for the future. It is a present reality that is transforming our institutions, social relationships, and educational systems. However, this progress must not come at the expense of the human rights or equity. From the Mexican perspective, building a strong AI ecosystem involves three key pieces. We have three pillars, our rights based on regulatory framework of substantial digital inclusion and national and international cooperation. The first one, regulatory development. Mexico does not yet have a specific law of artificial intelligence. Nevertheless, we do have fundamental instruments such as the federal law of the protection of personal data and the general law of the rights of children and adolescents, which establish clear obligation to protect the privacy, dignity and personal data of minors in the face of the technological use. Actually, in this vulnerable technological group, I will focus on this occasion. Tomorrow, on June 25th, a key point of agreement will be formally presented before the Permanent Commission of the Mexican Congress. It urges several government institutions, including the Ministry of the Interior, the Ministry of the Public Education and the Attorney General’s Office to work in coordination with local governments to design and promote comprehensive public policies and to protect children and adolescents from the risks posed by information and communication technologies. This initiative proposes that such actions go beyond just preventing cyberbullying or digital exploitation. It aims to promote critical digital literacy, foster a culture of cybersecurity and develop innovative legal frameworks such as the utilization of smart contracts and blockchain technology for the protection of sensitive data. The second key is the digital inclusion. An AI ecosystem must be built on the foundation of equity. In Mexico, the digital divide remains deep. Therefore, public policies must accompany technological development with universal access to connectivity, teacher training, and design of tools that respond to local needs. This is not merely a state obligation, this is an opportunity to create an AI with Latin America identity incorporating our languages, cultures, social challenges, and community aspirations. And the third, most important, the international cooperation. Mexico recognizes that AI demands a global approach, that is why we value spaces like this forum, where we can share common principles, ethical frameworks, and legislative experience, such as the European Union Artificial Intelligence Act, which has set a clear path for risk based on regulation. Mexican academia is committed to actively participating in these multilateral spaces to strengthen international cooperation and build governance framework that prioritizes the transparency, data protection, and accountability. Finally, in Mexico, it’s advising and building an ecosystem AI innovation and cooperation, driving by clear vision to promote technological development while upholding human dignity. The proposal to be presented tomorrow in the Congress reflects this commitment in this not enough to regulate after harm is done. We must anticipate, legislate, educate, and cooperate now. Only through shared responsibility can we build an ethical, inclusive and transformative artificial intelligence. Thank you very much.

Chin Yik Chan: Thank you very much, Ricardo. So you have brought here the three pillars from the Mexican perspective. And our next speaker will be Professor Dai-Li Na from the Shanghai Academy of Social Sciences.

Participant: Good afternoon, dear delegates and participants. It’s a great honor for me to have the opportunity to introduce the practical experience of Shanghai in building an innovative cooperation ecosystem for AI. My presentation will consist of two parts. The first part is Shanghai’s status and achievements in AI development. Shanghai is one of the largest cities in China. As a pioneer in reform and opening up, and a forerunner in innovation, Shanghai has long been committed to building itself into a global influential hub for scientific and technological innovation. Nowadays, Shanghai has set for itself a new goal, namely to establish itself as a global AI hub with a world-class AI industrial cluster, designating AI as one of its three leading industries. Shanghai has leveraged its comprehensive advantage, namely rich educational and scientific resources, a rebounding industrial foundation, abundant data resources, and a high level of openness to emerge as a leading city in China’s AI development. By fostering the coordinated growth of core AI technologies, fundamental software and hardware, intelligent products, and industrial applications, Shanghai has built a relatively complete AI industrial chain. The second part is pathways to build Shanghai’s AI innovation ecosystem. There are five key pathways. The first one is strengthening top-level design and leveraging the government’s guidance role. Shanghai has introduced a number of policies and measures to promote AI development. Key initiatives include these eight important policies. Now, the innovation ecosystem for fundamental models has flourished marked by projects like Xu Hui’s Model Speed Space and Pudong’s Model Power Community. The second is improving the AI industrial chain and innovation collaboration ecosystem. We have strengthened the local 4PLUSX innovation hubs by connecting leading enterprises and institutions in a fundamental model. Shanghai has fostered resourcing, sharing, and mutual promotion among dispersed entities, creating a differential innovation landscape. encompasses fundamental models in Xuhui River area, smart chips in Lingang New Area, humanoid robots in Zhangjiang, and intelligence robots in Maqiao. The third is deepening industry academia research integration to accelerate AI technology innovation. In talent development, Shanghai leverages rich academic resources to cultivate interdisciplinary professional. Examples are the AI colleges in Shanghai Jiao Tong Universities and Fudan Universities and Shanghai University of Electric Power. The fourth is driving AI innovation through industrial integration ecosystem development. Shanghai is advancing AI plus applications in six key sectors, finance, manufacturing, education, healthcare, culture, tourism, and urban governance. Last but not least is deepening international cooperation and activity participating in global governance. In international collaboration, Shanghai has done the following. It facilitates the establishment of headquarters on new ventures by global AI leaders, offering policy support in financing, research and development, talent, and markets. It leveraged events like China international import export and world AI conference to host the high-level academic events and connect in event with institutions providing an international cooperation platform for stakeholders to engage in AI industry development and governance. In addition, it engaged in global governance through institutions like National Test Zone Governance Working Group and the Shanghai Industry Security Committee while organizing forums on AI ethics, safety, and governance initiatives like Shanghai Initiative on AI Safety Development and Shanghai Declaration on Global AI Governance along AI development with the UN’s Sustainable Development Goals, ensuring human-centric and benevolent progress. Above all is all the five key pathways and ongoing forward. Shanghai will uphold its pioneering spirit to build a more dynamic AI innovation and collaboration ecosystem and contribute its strength to the global AI development and governance. Thank you for your attention and welcome to Shanghai. Welcome to China. Enjoy our smart city.

Chin Yik Chan: Thank you, Professor Depp for the introduction of the Shanghai experience. And finally, we are going to have a Wushu Yan. from China Mobile Research Institute.

Wushu Yan: Good afternoon, ladies and gentlemen. It is my great pleasure to attend this forum. The theme of my speech is harnessing wisdom and strength, working together to build an AI collaboration ecosystem. AI has become a driving force in technology and industrial innovation, transforming the production, lifestyle, and governance. Firstly, AI is being widely applied across industries. It is being embedded into industrial systems, enhancing the efficiency of value chain and creating new production from such as digital twins. Secondly, AI is reshaping lifestyles. It is redefining consumer experiences, employing artistic creation, and AI in smart devices boosts intelligence capabilities. Thirdly, AI optimizing governance. It enhances policymaking, improving public service delivery and fostering innovative approaches to public service. As AI evolves rapidly, we should address new challenges such as data privacy leak, algorithm bias, and the digital divide. Due to the cross-border application of AI technologies, the global flow of data, and the differences in legal regulations among countries, these changes are inherently global in nature. Therefore, we should strengthen technological exchange and cooperation, build global consensus on AI governance, and work toward an inclusive collaboration and open international ecosystem for AI. China has been an active supporter in building the global AI ecosystem. We are committed to promoting international exchanges in AI and participating in. and global AI governance. In recent years, China has put forward initiatives such as Global AI Governance Initiative, participating in Paris AI Action Submit and signed a statement on inclusive and sustainable AI for people and the planet. China has always put people first and championed the use of AI for good. We emphasize mutual respect, advocate for openness and sharing, and work to establish ethical standards for AI. Under the framework of the United Nations, we are committed to promoting AI that is inclusive, equitable, and sustainable for all. As a world-famous China operator, China Mobile has achieved global first in several aspects, including network skills, customer base, and revenue size. To seize AI development opportunities, China Mobile is implementing the AI Plus Action Plan and working as AI provider, aggregator, and operator. It has fully developed its strengths in three key areas, large-scale computing cluster, large-volume operator, and large-language model. For example, it launched two huge computing clusters and developed our old JiuTian Foundation model. Internally, AI is deeply empowering all business areas, significantly enhancing work efficiency and customer satisfaction. Externally, our AI solutions have already powered more than 1,000 real-world applications across health, energy, and other key industries. Additionally, China Mobile, along with partners, launched the AI Plus Pro Global Acquisition Alliance to promote our global AI adoption. Currently, it has already launched the AI International Large-Language Model Integrated Server and AI Plus Global Solutions.

Chin Yik Chan: Can I cut you off?

Wushu Yan: Our AI is driving a new wave of technological revolution and industrial transformation. It is reshaping the global economic society and even the human civilization.

Chin Yik Chan: Ms. Wu, I have to cut you off here. Thank you very much.

Wushu Yan: Thank you.

Chin Yik Chan: So let’s try to wrap up a little bit within this one minute. I end with perhaps one quote for Wolfgang. I think for Yuan, point out that AI is not a commodity and for Wolfgang, the UN should play a leading role in AI governance. And for Luigi, talks about the EU experience and for Dawei, ISOC and ChinaJF is very well connected. And Sajid, the sentence is no one stakeholder has all the answers. Ricardo, you emphasize human rights and equity. And for Delina, she talked about Shanghai experience and for our last speaker, he talked about some China initiatives. With that, I think I close this open forum and thank our speakers for their wonderful ideas. Also, I thank the director, General Qi Xiaoxia, for making this happen and also our audience. Thank you very much. Thank you. Thank you.

Agreement points

Multi-stakeholder approach is essential for AI governance

Speakers

– Sajid Rahman
– Wolfgang Klauweiter
– Chin Yik Chan

Arguments

Multi-stakeholder framework bringing technical expertise, public interest, and global perspective is essential

Internet governance principles and multi-stakeholder model can be applied to AI governance without reinventing the wheel

Multi-stakeholder forums like IGF provide essential platforms for AI governance discussions

Summary

All three speakers emphasize that effective AI governance requires bringing together diverse stakeholders including technical experts, governments, civil society, and international organizations, building on the successful multi-stakeholder model used in Internet governance.

Topics

Legal and regulatory | Development | Interdisciplinary approaches

UN should play a central role in global AI governance

Speakers

– Wolfgang Klauweiter
– Qi Xiaoxia

Arguments

UN should play leading role in AI governance as only platform including all countries

China supports Global AI Governance Initiative and participates in international AI governance efforts

Summary

Both speakers advocate for the United Nations to take a leading role in AI governance, with Wolfgang emphasizing the UN’s inclusivity and Qi Xiaoxia highlighting China’s support for UN-based initiatives.

Topics

Legal and regulatory | Development | Human rights principles

International cooperation is crucial for addressing AI challenges

Speakers

– Qi Xiaoxia
– Dai Wei
– Participant
– Ricardo Pelayo

Arguments

China supports Global AI Governance Initiative and participates in international AI governance efforts

Internet Society of China cooperates with over 20 global organizations including IGF, ICANN, UNESCO

Public-private partnerships are imperative to accelerate development and translate ideas into real-world impact

AI development must prioritize human rights, equity, and not come at expense of human dignity

Summary

Multiple speakers emphasize that AI governance challenges are global in nature and require extensive international cooperation, partnerships, and coordination across borders and sectors.

Topics

Development | Legal and regulatory | Human rights principles

AI development must be balanced with ethical considerations and human rights protection

Speakers

– Qi Xiaoxia
– Ricardo Pelayo
– Sajid Rahman

Arguments

China released first global legislative effort with interim measures for generative AI services management

AI development must prioritize human rights, equity, and not come at expense of human dignity

Digital divide concerns require ensuring equal access and preventing AI from deepening inequality

Summary

Speakers agree that while AI development should be promoted, it must be accompanied by strong regulatory frameworks and ethical safeguards to protect human rights and prevent negative societal impacts.

Topics

Human rights principles | Legal and regulatory | Development

Similar viewpoints

Both speakers recognize the rapid democratization and acceleration of AI technology, though they emphasize different aspects – Kurbalija focuses on accessibility while Rahman emphasizes unprecedented growth speed.

Speakers

– Jovan Kurbalija
– Sajid Rahman

Arguments

AI has become a commodity accessible to everyone, no longer mysterious technology limited to top labs

AI growth and transformation is unprecedented compared to previous technological waves

Topics

Economic | Development | Future of work

Both speakers advocate for applying proven Internet governance approaches to AI governance, emphasizing that existing multi-stakeholder frameworks provide a solid foundation for AI governance rather than starting from scratch.

Speakers

– Wolfgang Klauweiter
– Sajid Rahman

Arguments

Internet governance principles and multi-stakeholder model can be applied to AI governance without reinventing the wheel

Multi-stakeholder framework bringing technical expertise, public interest, and global perspective is essential

Topics

Legal and regulatory | Digital standards | Development

Both speakers present comprehensive, systematic approaches to AI implementation in China, emphasizing coordinated development across multiple sectors and the integration of AI into existing industrial and technological infrastructure.

Speakers

– Participant
– Wushu Yan

Arguments

Shanghai has built comprehensive AI industrial chain with coordinated growth across technologies and applications

China Mobile implements AI Plus Action Plan serving as AI provider, aggregator, and operator with global solutions

Topics

Economic | Development | Telecommunications infrastructure

Unexpected consensus

Standardization and open collaboration in AI development

Speakers

– Jovan Kurbalija
– Wolfgang Klauweiter

Arguments

Standardization around AI weights is crucial for international cooperation and avoiding social media platform mistakes

Internet governance principles and multi-stakeholder model can be applied to AI governance without reinventing the wheel

Explanation

Despite coming from different perspectives (Kurbalija from technical/innovation angle, Wolfgang from legal/regulatory angle), both speakers converge on the importance of building on existing standards and frameworks rather than creating isolated systems. This consensus is unexpected because it bridges technical and regulatory viewpoints.

Topics

Digital standards | Legal and regulatory | Development

Knowledge preservation and human-centric AI development

Speakers

– Jovan Kurbalija
– Ricardo Pelayo

Arguments

Knowledge preservation and bottom-up AI development should be prioritized over big platform dependence

AI development must prioritize human rights, equity, and not come at expense of human dignity

Explanation

Kurbalija’s technical focus on knowledge preservation and Ricardo’s legal focus on human rights unexpectedly align on the principle that AI development should serve human interests rather than concentrate power in large platforms or compromise human dignity.

Topics

Human rights principles | Cultural diversity | Development

Overall assessment

Summary

The speakers demonstrated strong consensus on several key principles: the necessity of multi-stakeholder approaches, the importance of international cooperation, the need for ethical AI development that protects human rights, and the value of building on existing governance frameworks rather than starting from scratch. There was also agreement on the UN’s central role in global AI governance and the need to prevent AI from exacerbating digital divides.

Consensus level

High level of consensus on fundamental principles with complementary rather than conflicting perspectives. The agreement spans technical, regulatory, and policy dimensions, suggesting a mature understanding of AI governance challenges. This consensus provides a strong foundation for collaborative international AI governance efforts, though implementation details and specific regulatory approaches may still require further discussion and negotiation.

Different viewpoints

Role of regulation vs. market freedom in AI governance

Speakers

– Wolfgang Klauweiter
– Qi Xiaoxia

Arguments

Risk-based regulatory approaches are needed but implementation remains challenging

China released first global legislative effort with interim measures for generative AI services management

Summary

Wolfgang emphasizes the complexity and challenges of implementing AI regulation, noting even the EU’s complex system faces simplification needs, while Qi Xiaoxia presents China’s proactive legislative approach as a successful model for AI governance

Topics

Legal and regulatory | Development

Centralized vs. decentralized approach to AI development

Speakers

– Jovan Kurbalija
– Qi Xiaoxia

Arguments

Knowledge preservation and bottom-up AI development should be prioritized over big platform dependence

China leads globally with over 60% of world’s AI patents and 430+ generative AI service models registered

Summary

Kurbalija advocates for bottom-up, decentralized AI development to preserve human knowledge and avoid big platform dependence, while Qi Xiaoxia highlights China’s centralized success in AI development through large-scale platforms and government coordination

Topics

Development | Human rights principles | Economic

Unexpected differences

Pace and nature of AI commoditization

Speakers

– Jovan Kurbalija
– Sajid Rahman

Arguments

AI has become a commodity accessible to everyone, no longer mysterious technology limited to top labs

AI growth and transformation is unprecedented compared to previous technological waves

Explanation

While both acknowledge rapid AI development, Kurbalija frames AI as already commoditized and accessible, while Rahman emphasizes the unprecedented and accelerating nature of transformation that still requires careful stewardship. This represents different perspectives on how mature and accessible AI technology currently is

Topics

Development | Economic

Overall assessment

Summary

The discussion shows relatively low levels of direct disagreement, with most conflicts being subtle differences in emphasis and approach rather than fundamental opposition. Main areas of disagreement center on regulatory approaches (proactive legislation vs. flexible frameworks), development models (centralized vs. decentralized), and the current state of AI accessibility.

Disagreement level

Low to moderate disagreement level. The speakers generally share common goals of responsible AI development and international cooperation, but differ on implementation strategies and governance mechanisms. This suggests a constructive foundation for dialogue, though the subtle differences in approach could lead to significant policy divergences in practice.

Partial agreements

Similar viewpoints

Both speakers recognize the rapid democratization and acceleration of AI technology, though they emphasize different aspects – Kurbalija focuses on accessibility while Rahman emphasizes unprecedented growth speed.

Speakers

– Jovan Kurbalija
– Sajid Rahman

Arguments

AI has become a commodity accessible to everyone, no longer mysterious technology limited to top labs

AI growth and transformation is unprecedented compared to previous technological waves

Topics

Economic | Development | Future of work

Both speakers advocate for applying proven Internet governance approaches to AI governance, emphasizing that existing multi-stakeholder frameworks provide a solid foundation for AI governance rather than starting from scratch.

Speakers

– Wolfgang Klauweiter
– Sajid Rahman

Arguments

Internet governance principles and multi-stakeholder model can be applied to AI governance without reinventing the wheel

Multi-stakeholder framework bringing technical expertise, public interest, and global perspective is essential

Topics

Legal and regulatory | Digital standards | Development

Both speakers present comprehensive, systematic approaches to AI implementation in China, emphasizing coordinated development across multiple sectors and the integration of AI into existing industrial and technological infrastructure.

Speakers

– Participant
– Wushu Yan

Arguments

Shanghai has built comprehensive AI industrial chain with coordinated growth across technologies and applications

China Mobile implements AI Plus Action Plan serving as AI provider, aggregator, and operator with global solutions

Topics

Economic | Development | Telecommunications infrastructure

Key takeaways

AI has evolved from mysterious technology to accessible commodity, with China leading globally through 430+ registered AI models and 60% of world’s AI patents

Multi-stakeholder governance approach similar to internet governance is essential, as no single stakeholder has all the answers for AI regulation

UN should play the leading role in global AI governance as the only platform that includes all countries, avoiding exclusionary approaches

AI governance requires balancing innovation promotion with risk management, using risk-based regulatory frameworks while avoiding strangulation of development

International cooperation must go beyond national agendas since AI is global by nature – data flows, algorithms and applications don’t respect borders

Digital divide and inclusivity concerns are critical – AI development must ensure equal access and prevent deepening of existing inequalities

Practical implementation requires comprehensive ecosystems combining government guidance, industry-academia collaboration, and international cooperation

AI should serve sustainable development goals and be governed by human-centric principles prioritizing human rights and dignity

Resolutions and action items

Use SDGs to both guide AI development and govern AI systems implementation

Develop standardization around AI weights to enable international cooperation and interoperability

Establish UN open source AI models with countries contributing their national AI models as contributions

Strengthen ties between global industry organizations to share experiences and promote best practices

Present comprehensive public policy proposals to Mexican Congress (specific action for June 25th mentioned)

Promote AI Plus Global Acquisition Alliance for international AI adoption

Build consensus on international AI governance through multi-stakeholder participation

Unresolved issues

How to effectively implement complex risk-based regulatory systems without strangulating innovation

Lack of specific AI legislation in many countries like Mexico, creating regulatory gaps

Transparency of AI algorithms remains challenging from both ethical and regulatory perspectives

How to bridge the intelligence divide and ensure developing countries have equal representation in AI governance

Preventing technological protectionism while maintaining national security interests

Addressing data privacy, algorithmic bias, and digital divide challenges that are inherently global

Determining what specifically should be regulated in AI governance without hindering development

Suggested compromises

Risk-based regulatory approach with different levels of regulation based on AI application risk levels (forbidden for human dignity violations, strict rules for high-risk, flexible for low-risk, no regulation for no-risk)

Building on existing internet governance frameworks rather than reinventing governance mechanisms for AI

Combining open innovation with trust and transparency at the core to avoid technological protectionism

Balancing national AI strategies with international cooperation through shared forums and joint projects

Using holistic approaches that don’t exclude any stakeholder groups while maintaining efficiency

Promoting open source tools and accessible platforms while allowing for commercial development and monetization

AI is becoming a commodity. Only three years ago, AI was a mysterious technology in the hands of a few developers and the top labs with huge investment of money. That has changed… It’s a time to revisit and to see what is basically our governance should regulate and what should be the governance function that we should cover by governance models.

Speaker

Jovan Kurbalija

Reason

This comment fundamentally reframes the AI governance discussion by challenging the premise that AI remains an exclusive, high-barrier technology. It introduces a critical temporal perspective – that governance frameworks are based on outdated assumptions about AI’s accessibility and democratization. This insight forces a reconsideration of regulatory approaches.

Impact

This comment established a new analytical framework for the entire discussion. It shifted the conversation from traditional governance models to questioning what actually needs to be governed in an era of democratized AI. Subsequent speakers referenced this commoditization theme, and it influenced how participants discussed accessibility, regulation complexity, and the need for adaptive governance frameworks.

Knowledge is what defines us as a human beings. This is definition of our human dignity, our personal knowledge and knowledge of our communities… Somehow knowledge is cleaned from the political discourse of the last 20 years. You don’t have any more knowledge, you have only data.

Speaker

Jovan Kurbalija

Reason

This observation provides a profound philosophical critique of how the discourse around AI has shifted from knowledge preservation to data extraction. It connects AI governance to fundamental questions of human dignity and community identity, elevating the discussion beyond technical considerations to existential ones about what makes us human.

Impact

This comment introduced a humanistic dimension that influenced the tone of subsequent discussions. It provided intellectual grounding for arguments about bottom-up AI development and community empowerment, and can be seen reflected in later speakers’ emphasis on human rights, equity, and inclusive development.

The question is not whether we should cooperate. It is whether we do fast enough, broadly enough, and wisely enough. To steer AI towards humanity’s greatest aspiration, rather than deepest fears… We won’t have a second chance.

Speaker

Sajid Rahman

Reason

This comment reframes the urgency of AI governance from a technical challenge to an existential imperative. It introduces the concept of irreversibility and historical responsibility, suggesting that current decisions will have permanent consequences for humanity’s future.

Impact

This statement elevated the stakes of the entire discussion and provided moral urgency that influenced how other speakers framed their contributions. It shifted the conversation from incremental policy discussions to recognition of this as a defining moment for civilization, influencing the tone of subsequent speakers who emphasized comprehensive, immediate action.

There is no need to reinvent the wheel because we are on the Internet Governance Forum… AI governance or AI is just using the Internet. So that means you can build on what has been developed in the last 20 years in the Internet Governance Forum with the Internet and you can use the definition for Internet governance also for AI governance.

Speaker

Wolfgang Klauweiter

Reason

This comment provides crucial institutional continuity by connecting AI governance to established Internet governance frameworks. It challenges the assumption that AI requires entirely new governance structures and offers a practical pathway forward by leveraging existing multi-stakeholder models and institutional knowledge.

Impact

This insight provided a concrete foundation for subsequent discussions about multi-stakeholder approaches. It influenced how speakers like Sajid Rahman elaborated on multi-stakeholderism and gave legitimacy to applying proven governance models to AI challenges, reducing the perceived complexity of building new international frameworks.

We should have the standardization around the AI weights… We already went a step in this trap with social media, where when you are on one social media platform, you cannot bring your contacts on the other social media platform. Let us learn something from our mistakes.

Speaker

Jovan Kurbalija

Reason

This comment demonstrates strategic learning from previous technological transitions and identifies a specific technical intervention (weight standardization) that could prevent the creation of AI silos. It connects past governance failures to current opportunities for better outcomes.

Impact

This technical proposal provided a concrete example of how international cooperation could work in practice. It influenced the discussion by showing how governance principles could translate into specific technical standards, and several speakers subsequently referenced the importance of avoiding technological fragmentation and promoting interoperability.

Overall assessment

These key comments fundamentally shaped the discussion by establishing several critical frameworks: the democratization of AI requiring new governance approaches, the philosophical stakes involving human dignity and knowledge preservation, the existential urgency of getting AI governance right, the practical value of building on existing Internet governance models, and the importance of learning from past technological governance failures. Together, these insights elevated the conversation from routine policy discussions to a recognition of AI governance as a defining challenge for human civilization, while simultaneously providing practical pathways forward through established multi-stakeholder approaches and specific technical interventions. The comments created a productive tension between urgency and pragmatism that influenced how all subsequent speakers framed their contributions.

How can we standardize AI weights sharing across different platforms to enable international cooperation?

Speaker

Jovan Kurbalija

Explanation

This is crucial for preventing the same mistakes made with social media platform isolation and enabling true international AI collaboration through interoperable systems.

How can SDGs be practically implemented as governance frameworks for AI systems beyond just using AI to achieve SDGs?

Speaker

Jovan Kurbalija

Explanation

This represents a novel approach to AI governance using existing international consensus on human values and priorities as regulatory frameworks.

How can we preserve and protect community knowledge in the age of AI without losing it to big tech platforms?

Speaker

Jovan Kurbalija

Explanation

This addresses fundamental concerns about knowledge sovereignty and human dignity in AI development, particularly relevant as AI becomes commoditized.

What is the practical difference between Internet governance and AI governance, and how much can be built upon existing frameworks?

Speaker

Wolfgang Klauweiter

Explanation

This is important for determining whether entirely new governance structures are needed or if existing Internet governance models can be adapted for AI.

How can the complex European AI Act risk-based approach be simplified for practical implementation?

Speaker

Wolfgang Klauweiter

Explanation

The European experience shows that complex regulatory frameworks may need simplification, which has implications for other jurisdictions developing AI regulations.

How deep will the digital divide become with AI implementation, and what specific measures can prevent it?

Speaker

Sajid Rahman

Explanation

This builds on lessons from Internet 2.0 digital divide issues and is crucial for ensuring AI benefits are distributed equitably globally.

How can transparency of AI algorithms be ensured from both ethical and regulatory perspectives?

Speaker

Sajid Rahman

Explanation

This is becoming a critical challenge as AI systems become more complex and their decision-making processes less interpretable.

How can AI development incorporate local languages, cultures, and community aspirations, particularly in Latin America?

Speaker

Ricardo Pelayo

Explanation

This addresses the need for culturally appropriate AI that serves diverse global communities rather than being dominated by a few major languages and cultures.

How can the concept of ‘knowledge’ be reintegrated into AI policy discourse after being replaced by ‘data’ in recent years?

Speaker

Jovan Kurbalija

Explanation

This represents an important research area for understanding how policy language evolution affects AI governance approaches and human-centered development.