ECOSOC Special Meeting on Artificial Intelligence (AI)

How do we govern technologies if we don't yet know their full potential?

This rise, I would say explosion of generative AI is putting the thoughts how to regulate it, how to govern it in everyone's minds, in everyone's discussions, including policymakers.

AI technology is a promising technology to accelerate sustainable development. But on the other hand, we need to effort to mitigate the negative harms and negative sides.

If we're spending trillions of dollars on increasing AI capabilities, shouldn't you be spending 5% of that, like $50 billion, on getting all the governance— upgrading the governance itself?

We need to create institutional structures and we need to have proper policy framework.

Governance is surely broader than regulation, and therefore we need to reduce the fragmentation, whether it's amongst UN agencies or governments or regional bodies to ensure that we have standards that we can all comply to around AI governance.

The most important thing is actually to look at the incentives that we're training the AI systems to achieve.

We need to view governance not as an inhibitor, but as an enabler, an enabler For AI for good. For AI for good.

Hi, everybody. Delighted to be here. My name is Bob Rae. I'm the President of the Economic and Social Council. We're delighted to have all of you here for our special meeting on artificial intelligence. I'm going to be joined on the podium in a minute by Undersecretary Amandeep Singh Gill. And we're really looking forward to a lively, interactive discussion. As you can see, we've changed the makeup of the room. Our Economic and Social Councils don't normally look quite as exciting as this, but we're delighted to be able to take advantage of what's available. And really, the theme that we want to set out on this morning is, how can we make sure that artificial intelligence can be, can be a force for good? We're going to be having another meeting in Geneva in, in July, under the leadership of the ITU to talk about this theme as well. And we think that at ECOSOC, we have a responsibility to really try to encourage a conversation about, about AI and about how we can make it work for everybody, and how we can make sure that it's not just a, some sort of technological gimmick that everybody looks at with great fear, that ends up enriching some countries and some people, a form of technology that transforms, in many cases, the way we do things, but doesn't really help to deal with the overarching problems and challenges we face in the world today, the enormous inequality that exists, the gaps between those who have and those who don't have, the digital divide that separates countries that have access to technology and those parts of the world that don't. And so we're really looking at how we can build a greater knowledge of how to take advantage of AI for for the whole system. Of course, there's profound questions around ethics. There's important questions of recognizing that technologies in and of themselves don't have an inherent moral compass. And then we have to decide what to do about it. Can we allow AI— can we have a system of governance And also a use of AI that actually empowers people? Or is it simply going to accelerate the exploitation of people that has so far gone unchecked? There are people who think somehow that AI can be stopped. And I've got news for you. It can't be stopped. The notion that technological, technological change will— is something that we can control by passing a resolution at the United Nations? No, that's not going to happen. This is a life force that goes well beyond the boundaries of this institution or any other institution. And this year at the UN, like every year, and my predecessor had— we had a meeting in Santiago on these questions of trying to look at the relationship between AI and decent work. But we've had many different conferences and discussions that go beyond the boundaries of the UN. The Hiroshima AI process, the Global AI Summits convened at Bletchley Park and in Seoul and London, and the AI Action Summit in Paris. They have all highlighted the urgency of making sure we can forge common principles, shared safeguards, and interoperable standards to guide the responsible development of AI. The conference— the conversation that we're starting today under the auspices of ECOSOC will continue for the next several days at our Multi-Stakeholder Forum on Science, Technology and Innovation for the SDGs, and it will be an opportunity to carry on the dialogue that we have already started, but that we're going to be starting and continuing this week. The clear thing I just think is so important for us to say is that what AI is all about is not going to be determined by a handful of actors, no matter how powerful they may be. They're going to have to listen to the voices of the people and of their governments. When the Gutenberg printing press was invented a few hundred years ago, We didn't have a United Nations, and we didn't have an organized discussion about what should be done with this new technology. But suffice it to say that the printing press transformed the world. It transformed literacy. It transformed people's sense of how they could, how they could put themselves at the center of the conversation. Arguably, in the field of theology, in the field of politics, nothing was more momentous than the discovery of the printing press and of the books and the literacy to which it gave rise. I think we're now in the middle of an era when a new form of technology is going to have just as dramatic an impact as the printing press did. And yet we have the opportunity to perhaps be a little more knowledgeable about how we can unlock this potential. There are countries that have done it right. There are countries that have really tried to make a difference in giving more and more people access to the technology and used it for the benefit of the SDGs. And there are countries that have got it wrong. And so we need to discuss among ourselves how we can get it right and how we can make sure that this extraordinary breakthrough in technological understanding can be used Yes, for the public good and for the benefit of all mankind. So with that, I'm going to ask Mr. Gill to say a few words. We've all appreciated his leadership here in New York and we look forward to hearing his comments. Thank you.

Thank you very much, Ambassador Ray, Bob, Excellencies, ladies and gentlemen. This is an important moment. The STI Forum is meeting this week, and we have the special meeting of the ECOSOC on artificial intelligence. Our world is literally changing. Our known Landmarks are disappearing before our eyes, whether they are geopolitical, whether they are related to international trade, and technology is a big, big factor in this rapid change. Hardly a month, hardly a week passes without artificial intelligence rewriting what's possible. This morning we had the Action Day here in this room with this Nice stage, and one of the presenters mentioned that our policy cycles are moving at a sixty-month pace, five-year pace, and artificial intelligence is rewriting itself every six months. So this is our challenge, and the related challenge, and Bob spoke to this, is the risk of the AI. Divide. So we have a lot of existing divides. Is AI going to widen those divides, widen access to opportunity or not? That is one of the biggest questions facing us at the UN. I was recently at a meeting of economists, and one of the economists in the room was the chief economist of a leading AI lab, one of the top labs in the world. And the discussion there was on what is the productivity impact of AI. And the estimates today, the best estimates range from about 0.4% per annum to about 12% per annum. And just remember that even in advanced economies like the US, productivity growth in recent decades has never gone beyond 2.5%, 3%. Even in emerging economies, this is in the range of 5%, 7%. So the question in the room was, if you have that kind of a productivity growth, where are the benefits going to go to, and are governments going to be even able to tax some of those productivity gains, and who gets impacted? What is the impact on labour markets, etc.? So these are our challenges, and of course the opportunity is immense. You know, just 3 weeks ago, clinicians in Benin field-tested a mobile ultrasound system whose on-device AI diagnosed tuberculosis with a 93% sensitivity. Impressive. Not yet at that level of ultra-experts, but surpassing most human experts and needing no laboratory. So, this is just one milestone that we see on the opportunity side. And all of these challenges related to equity, access, the divides, the opportunities, and there are risks, plenty of risks, would demand governance. Would demand capacity, capacity related to policy in particular, would demand resources, financing. We are at a moment where the SDGs are not doing well, where we are struggling to find less constrained fiscal spaces, so this opportunity has to be taken seriously. The past year has shown that we at the UN can actually make progress on this issue. We can make a difference to the way AI develops, it's governed, it's used. As you know, last September, the GA adopted the Global Digital Compact as part of the Pact for the Future, the first universal agreement to peace, human rights, inclusion, and sustainability together and put them at the heart of the digital transformation. And in this area of AI, there are 3 concrete commitments taken by member states. The first one, let's call it evidence first. Member states have agreed to establish an independent international scientific panel on AI so that everyone can draw on a common rigorous evidence base. I see one of the co-facilitators of the modalities of that process is in the room. So this is an important decision that needs to be actualized, operationalized this year. The second area of action is rules that work together. So member states need to convene a global dialogue on AI governance. To share best practices, improve interoperability of rules, and build capacity of countries to design, deploy, and oversee AI that advances every Sustainable Development Goal— capacity related to governance through mutual learning. We are not yet speaking of top-down normative approaches, but how best— how to best find the articulation between soft norms, soft governance, and what we already have at the United Nations. The third area of action is financing that leaves no one behind. Financing for AI capacity building, financing to advance the critical enablers that go into building and deploying AI, whether it be human talent, datasets, or the compute infrastructure that's so critical. I was in Africa recently. All of Africa put together has less than 1,000 GPUs, graphics processing units, critical for training AI models. And just a single US company, Meta, has more than 250,000 GPUs. So it just tells you the scale of that problem. So the SG has been asked by member states to craft a report with innovative voluntary funding options, drawing on the High-Level Advisory Body on AI's proposal for a global fund on AI to present it to the General Assembly at its 79th session. So, these resources hopefully will help us build digital public goods and AI capacity with special attention to the Global South, and hopefully would help us advance the Sustainable Development goals. Now, the world is not standing still. I mean, we are moving at the UN, but the rest of the world is, particularly in the regions I mentioned— I was in Africa— the African Union's Continental AI Strategy, the ASEAN's Guide on AI Governance, the Latin American coordinated approach through the Santiago Declaration, All of these developments tell us here in New York that we need to move forward with urgency on these 3 global pathways that come out of the GDC, but we also need to work closely with the regions because the regions are dynamic, they are taking their own steps, and in a future where the UN has to be even more fit for purpose, I think there's no alternative to networked multilateralism. We advance the global processes, but we also do that in collaboration with these regional frameworks, which demonstrate a crucial shift from being— from these regions being recipients of governance models to becoming architects of inclusive approaches that balance innovation, with equity and with risk management. So, in closing, distinguished delegates, Excellencies, ladies and gentlemen, the GDC, the Global Digital Compact, provides us with a framework, but frameworks alone do not create change. The task now before us is to translate that framework, those principles, into practice, especially by following through on these 3 pathways and by taking coordinated action with regions, with dynamic regions on these. Thank you very much, and I'm grateful for this opportunity to share some thoughts.

Prioritized in the Global Digital Compact and gaining momentum as one of the UN Secretary-General's High Impact Initiatives, DPI is delivering impact at scale. It is turning possibilities into realities and empowering countries to transform our planet and people's lives all the way to the last mile. From Malawi to Finland to Brazil, DPI is a transformative people-first movement. This movement recognizes the need for safeguards as an upfront necessity, not an afterthought. Embedding universal principles across every step of the DPI design and implementation journey ensures that the rights of people are respected and preserved and our planet is protected. This includes in Malawi, where 97% of Malawians have a legal digital ID being used to access services. This adoption saved the government $43 million in annual payments that were previously made to ghost beneficiaries. The same in Finland, where the X-Road data ecosystem offers over 800 public digital services and has scaled as a public good around the world. With a commitment to safeguards in the Brazilian Digital Government Strategy, the Rural Environment Registry is helping to achieve the goal of the Paris Agreement by reforesting 12 million hectares and creating interoperability between stakeholders and infrastructure across natural resources. But it is not just Malawi, Finland, and Brazil. The impact of DPI is growing, with more and more countries leveraging the transformative power of digital public infrastructure. To keep the promise of the Sustainable Development Goals, DPI represents a fundamental pivot that involves everyone, everywhere, and that includes you. Let's work together to build a safe and inclusive digital future for all.

Around the world, corporate leaders are grappling with a new and pressing priority: the adoption of artificial intelligence across their businesses. McKinsey calculates up to 72% of organizations globally already use AI in their operations. And 91% of C-suite executives polled by Thomson Reuters say they have plans to adopt AI tools within the next 18 months. Corporate leaders are navigating the difficult task of embracing the potential of AI, fending off competition, driving innovation, and managing new risks. To do so, they must consider critical factors: transparency, governance and oversight, changing legislation, the impact on workers and democracy. The challenge is immense. So how can companies navigate this rapidly evolving landscape? This is why we are launching the AI Governance Disclosure Initiative, produced by UNESCO and the Thomson Reuters Foundation. This survey will help companies map their AI adoption journey across operations, products, and services. The AI Governance Disclosure Initiative will act as a North Star for CEOs, CTOs, corporate councils, and board directors. The survey covers issues such as governance, procurement, auditing, environmental impact, human oversight, data privacy, bias mitigating measures, and the impact on workforce. The AI Governance Disclosure Initiative builds on the recommendation of the Ethics of AI, a framework developed by UNESCO and applying to all its 194 member states. Along with the Thomson Reuters Foundation's expertise in advancing responsible business practices. A first of its kind, the initiative can play a critical role in helping corporate leaders mitigate AI risks and ensure new technologies are deployed in ways that are ethical and do no harm. Find out more about how your company can get involved and join the initiative, positioning your organization at the forefront of transformational change.

Okay, welcome everyone. Good afternoon. No? No one? Good afternoon. All right. Let's wake up. This is AI we're talking about. This is dynamic. This is agile. My name is Robert Opp. I'm the Chief Digital Officer for the United Nations Development Programme, and it's my distinct honor to be the fireside chatter, the moderator of this fireside chat with 2 incredible speakers. And I just want to say I was given permission by the president of ECOSOC to keep this informal. So I'm going to say all protocols observed now, and I'm going to do away with the Excellencies and distinguished so-ons and so-forths. So we have heard a the opening speakers of this session talk about the issues related to AI that we're facing today, and things are developing at a dizzying speed. Of course, AI is poised to virtually change every aspect of society as we know it, potentially. And that opens huge opportunities and potentials for achieving the Sustainable Development Goals, addressing some of those longstanding development challenges that we've had, and endure with us and looking at a bright future of humanity. And yet we know there are challenges, big challenges. Challenges like access, equity, the risks involved, many other things that we're going to get into. And the good news is that I have a couple of speakers who know a lot about this and are going to help us work our way through it. So, this fireside chat is really going to look at these issues, the opportunities and the challenges, and really think through what are some of the things we need to do and the things we need to decide as people around the world about what we want AI to do for us. So, I am going to do a quick introduction of our speakers and then go straight into the discussion. And like I said, I have 2 incredible people here who I've already— who I've worked with before on a number of occasions, who I respect deeply. And our first speaker I want to introduce is His Excellency Philip Thigo, who's the Special Envoy on Technology at the Office of the President of Kenya, Kenya's tech ambassador. Philip is a leading voice in this space. He was part of the High-Level Advisory Board on AI that the Secretary-General put together. He's the founding director for Africa at the Thunderbird School of Global Management. He's worked across different foundations, different international organizations, and many, many other things. He's virtually everywhere at every time. The other— our other speaker is Dr. Rachel Adams, who is the CEO of the Global Center on AI Governance. And Rachel brings a deep background of human rights and AI in the spaces and has advised many different institutions from UNESCO, who we just saw, Gates Foundation, and many others. She also has honored us by being an advisor on some of our projects recently as well, including some AI initiatives. She has got research roles at the University of Cambridge and the University of Cape Town, And she directs the African Observatory on Responsible AI. And she's behind the Global Index on Responsible AI, which has helped us, as UNDP, do a lot of— inform our work when it comes to looking at the kinds of things, measures that countries can take in ensuring responsible use of AI. Okay. So, let's get started. Rachel, I'll start with you. One of the points that was made in the opening remarks was around the AI divide. And the divide is staggering. And you've just released a new book. I think it came out in January. I'm partway through it, haven't finished it yet. But you make the point about the current projections around economic growth and AI and how the United States, China, to a certain extent, Europe and others are poised to have huge growth. And the tech envoy was just mentioning this as well. But Africa and other parts of the world are expected to do, in the current projections, have much, much less portion of that growth. And we know that only around 5%, or rather 2%, if I'm not mistaken, 2% of the world's data centers are based in Africa. And only 5% of AI innovators in Africa have the compute power that they feel they need. So when we look at this AI divide, how can we ensure that the development and deployment of AI technologies don't entrench existing inequalities, which is a lot of what your book is about, but instead become a force for equity?

Thank you so much, and it's such a pleasure to be here with you all today. I think it's important to start off by recognizing that AI in and of itself is not a force for equity. Its harms disproportionately affect women, people of color, underrepresented groups. As we know, to be able to access and build these technologies requires huge resources and a lot of capital and money. And then we also know that the effectiveness of these systems depends on their deployment in effectively already sophisticated digital environments that we don't have across much of Africa and certainly much of the majority world. world. So my book was an attempt to really try and articulate this very stark divide and think about what are the efforts that we need to take to address it, many of which I think have been well laid out in the opening today. So I think really important here is international cooperation, and we need it at a scale that we've perhaps never imagined before. I can think of the Artificial Intelligence for Development program that is the longest-running and largest funding program on AI in Africa, supported by the UK and Canadian governments. And the kind of things that they're investing in is equitable access to compute resources in Africa and also supporting local innovators, which is really, really important to build language-based technologies in local African languages, potentially bringing the benefits of AI to millions more people. I'll pause there. Thank you.

No, thanks, Rachel. And you mentioned Canada's International Development Research Centre has done incredible work in this space, and I think has supported some of the work that helped set your organization up and things. I also just want to mention, I mentioned your book, but I didn't mention the name. It's called The New Empire of AI. And I, as I mentioned, I'm partway through, but it's really, really interesting and explores these issues in depth.

Thank you.

So, congratulations on the launch, Rachel. Philip, I'll turn to you, and I want to kind of talk about this from an ecosystem point of view. We know that innovation and technologies happen in ecosystems, and we know that those ecosystems in Africa in particular are seriously challenged. So, whether it's Africa or elsewhere in the Global Majority, how do you see AI-driven innovations really being able to scale or replicate at scale in a much more serious way or a faster way than we're seeing currently?

Thank you so much, Robyn. Always a pleasure to be with Rachel on the same panel. So, I look at it in this perspective. One, I think AI is not new, and there's an assumption that people think it's new. It's been there from the 1950s. It's just that this iteration, this version of it, just started being popular in 2022. So, A lot of our innovation ecosystems was there even before this kind of version of AI. And so I think it's always a matter of perspective, and that's why sometimes I really talk about fluency. It's not to put a lot of pressure on the global majority to build the same way the Global North has built its ecosystem. And so I think if you look at Africa, for example, we have more than 500 innovation hubs and labs that are vibrant. We have about 300,000 Kenyans and a million Nigerians on GitHub. So, these ecosystems and this talent is there. The idea for me is how do you harness it for the development of the continent? And we know the previous revolutions basically have been built on the back of Africans. I spend my time in Silicon Valley, and a lot of the researchers who are there are either Kenyans, Nigerians, Ghanaians, South Africans. And so, I think for me, the potential of scale is about recognizing that the continent in one way or another is engaged on AI. What we need to flip the model is their engagement needs to benefit the continent, and for that to happen, we need to invest in what matters, invest in connectivity so that it drives innovation, not connectivity just for other things. We know a lot of Africans spend a lot of time on social media, so sometimes it's not about the connectivity, it's the use of it. How do we build programs and capabilities that people can leverage so that they can learn? The second part, of course, is energy. As you know, the continent is still not connected very well, so how do we ensure that we have that kind of access? The other part, and Rachel mentioned it a bit and you mentioned a bit, is compute. So I spend a lot of time, for example, with Strive and Kasava Technologies around ensuring that compute is equitable, because it doesn't make sense for an African innovator to pay the same as somebody in Silicon Valley or somebody in London. And so how do we ensure that there's not just federated, but there's a model within which the cost makes sense?

No, absolutely. And, you know, you've both been kind enough to be involved in a project that we have been working on with the government of Italy on creating an AI hub for sustainable development in Africa. And compute is one of those parameters that we look at. And rather than, you know, forget looking at affordability, at the moment, African innovators pay much more than you do in the United States or Europe to access your compute power. with greater latency and more connection problems. So, it is definitely— we see that as a bottleneck as well. If I go back to you, Rachel, and look at this issue of, which we started to get into, being more inclusive, responsive to diverse cultural, socioeconomic contexts, how do we make AI governance frameworks more inclusive? And how do we avoid those fueling inequalities and fueling and entrenching imbalances of power, the kinds of things that you mentioned in your book? How do we avoid those by strengthening AI governance frameworks?

Thank you. Well, I think I was very involved with the African Union's continental AI strategy, and I think they took an important step in centering human rights in their AI governance approach. Human rights from social, cultural, economic, and labor rights remains, in my view, the most important touchstone to ensuring that AI governance is just and inclusive. So our study, the Global Index on Responsible AI, which measured rights-respecting AI governance in over 138 countries, including 42 in Africa, found that the majority of countries around the world were not taking steps to ensure that human rights were being promoted and protected in the age of AI. AI. We're actually very happy to announce that the Global Index is being funded for a second edition, again by the UK and Canadian governments, to be published in, in 2026, which will give us really important insights into the global state of responsible and rights-respecting AI, but also allow us to measure progress over time. I also think that as we are investing more in building AI capabilities, as we are seeing these systems become more complex, as we are adopting them more widely, the need to safeguard them from nefarious and unauthorized use becomes even more urgent. Where AI poses large risks, serious risks to disrupt political stability, to undermine democracy, to violate human rights at scale, the power and value of these technologies to advance progress towards the Sustainable Development Goals becomes diminished and threatened. And this includes the use of new AI technologies, like generative AI, to produce and disseminate synthetic, intimate, non-consensual imagery at scale, disproportionately affecting the safety and security of women and girls. So, we need concrete measures to address these abuses. We need globally inclusive early warning systems and information-sharing systems around AI safety, trust, and security. We need to ensure that gender-based tech-facilitated gender-based violence is also an issue of AI safety, because if we don't do this, we really are going to deepen those inequalities that I think we're all so concerned about.

Can I ask you just a follow-up? And talking about the Index for Responsible AI, you know, I think you've mentioned a few of those things, but You know, we have a lot of member state representatives here, permanent mission members and things. What goes into the index itself?

Thank you so much. So, in the first edition, we had 19 thematic areas of responsible AI, which covered from public procurement to children's rights to the standard kind of AI ethics principles, proportionality and do no harm, transparency and explainability, accountability, and so forth. And then we measured— we had an expert survey that was completed by country experts in every single country. So, we created this amazing global network of responsible AI researchers. And they assessed what government frameworks exist to address this issue, say, children's rights in AI, what steps have governments taken to implement those frameworks, and in the absence of either of those, were there non-state actors such as universities and civil society organizations that were taking steps steps to promote and address these issues in the absence of government efforts. We found that particularly among the kind of inclusion and human rights-related topics like children, women and gender, like cultural and linguistic diversity, universities and civil society organizations are really far outweighing what their governments were doing to promote these topics.

No, thanks, Rachel. And the reason I ask partly is that, you know, a lot of the index gives an interesting basis or entry point for us when we are often requested by countries to look at their AI ecosystems or AI strategy or build an AI roadmap of some kind. We start with the index, and that's really helpful to kind of look across those different parameters. What do countries have in place? What's not in place? What can they do more? and that sort of thing. So thank you. Okay, I'm going to turn to Philip for one more question, but I want to warn you, it's— after that, I'm going to open to the floor. So you can indicate if you want to take the floor, and then I'm going to try to track that, and we'll go to interventions from people in the audience, if it all works well and if technology does not fail. So Philip, my last immediate question for you is around some of the issues that you and I have talked about a lot already, which is, you know, trust and so on. And I should mention, Rachel's not the only author here on stage, because Philip recently authored an article that came out in the Brookings Institution called Shaping Africa's Inclusive and Trustworthy Digital Future. And in that article, Philip, one of the lines that really stuck out to me was, Africa must use— sorry, Africa must move from being merely a user to a creator of technology. How do we do that? And how do we prioritize trust, inclusion, ethics at the same time?

Thank you so much, Obam. I think Rachel has mentioned some of these elements, of which I say quite critical. And because every time we talk about AI, we talk about safety, we talk about security, we talk about inclusivity, but what we do not talk about is trust. And I think for me, trust must underpin everything around AI. And if it's about trust, then it means the communities that are impacted or affected by AI must be part of its creation and building up of that technology. And so, for me, it's not just about Africa must just be co-creators. I think there has to be an imperative. And for me, the ethical piece must mean that Africans must be involved in how AI is built or designed. Otherwise, we'll continue to hear about bias without having a way to shift that bias. The other piece, of course, is about partnerships that are more equitable and more transparent, because currently AI is concentrated amongst a few countries and a few companies. And part of building that trust is to ensure that everybody is part of the development of AI if it's going to impact everything you know, everything that we do. I think the third bit for me, and this is what I see, sometimes people say, but the continent is not ready. But my response is that why can't we have low-barrier use cases that are more applicable, that are more nuanced, that more speak to our reality? And I think there's possibility for that. Rachel has mentioned matters of linguistic diversity, also geographic diversity. And I think for me, part of it is also to ensure that AI speaks to us. Speaks to the over 2,000 languages and dialects in the continent. But also, I see that as an economic opportunity. So, I think for me, even for business, it makes good business sense to work with continents like ours, Latin America and Asia, to be able to build technologies that are more nuanced and speak to their realities, but also solves those local problems. And that's— and part of that is building trust.

Yes, and both of you have been helping us understand more about this notion of trust and what it means in a context like Africa or the global majority. Because trust and safety as a concept really came about from tech companies to start with, and partly about, you know, trust with consumers and partly about reputational risk and things. But as we take these discussions into, you know, African contexts and things, we really have to think, what does that mean? What's different about it? And how do we establish the trust with people. Okay, well, thank you both for those initial questions. Now, as I said, we are going to try to see if technology will work and tell me by opening my iPad if I have a speakers list here. And I do. It's magic. So, I will call on the member states and others who've asked for the floor. We'll probably group these a little bit, and I might go back and forth a bit with the panel for their comments and reflections and so on. And just as a warning, the sort of regular rules kind of apply here. If you're speaking as an individual country or an individual organization, you'll have 2.5 minutes. And if you're as a group, you'll have 3 minutes. And unfortunately, the mics cut off at those points. So, apologies in advance for those of you who actually end up getting cut off. All right. The first speaker on the list is Malta.

Thank you, Robert. And actually, your last question about trust is a good segue because my intervention was about trust in AI. Pioneering our 2019 strategy document, Malta Towards AI: Malta's Ethical AI Framework, Malta advocated a sustainable environmentally and environmentally friendly AI as one of the fundamental requirements to achieve trustworthy AI. Sustainable and environmentally friendly AI requires first mechanisms to measure the environmental impacts across its deployment and use. Second, actions to minimize negative impacts throughout the system's lifecycle. Malta is currently realigning its national AI strategy to address the Sustainable Development Goals. Malta's Vision 2050, launched in 2024, is a comprehensive 25-year plan that anticipates the sustainable use of AI to enhance the quality of life, particularly in key sectors like digital infrastructure, education, and healthcare. The Maltese government will continue supporting businesses in adopting AI solutions that reduce environmental impacts. As noted in the Digital Diplomacy Profile 2024, Malta also acknowledges AI's role in diplomacy and foreign policy. Malta has actively promoted digital diplomacy to the Global South through its founding and ongoing support of the Diplo Foundation, a leading academy on digital topics, including AI. Through the Malta Scholarship Fund, Malta will continue to be instrumental in the digital training of diplomats and government officials from small and developing countries. AI technologies used maliciously, such as deepfakes and voice cloning, are eroding trust in what we see and hear by enabling the faster, cheaper creation of false content. They also contribute to the fragmentation of the public sphere by reinforcing biased views through personalized filters. Conversely, AI can aid efforts to counter such information manipulation by detecting unethical behavior, analyzing content and user profiles, and tracking the spread of narratives across social networks. Countering disinformation requires transparency, effective governance, and empowering citizens, especially vulnerable groups, through education, media, and

Thank you so much, Malta.

Okay, our next speaker is South Africa. I think I saw Ambassador here.

Thank you very much, and I really wish to thank you for that interaction. I found it very useful and very insightful, so I take on board some of the thoughts that you have shared with us. It is South Africa's firm view that the following should happen. To maximize AI-driven solutions to positively impact on SDGs. Firstly, appreciate the correlation between STI and AI, and this appreciation should be born from the simple maxim that, much like STI, AI offers the innovative means of accelerating implementations of SDGs. And to this end, national legislative efforts should work to enhance cross-pollination between STI and AI. Second, building on that correlation, during the 8th STI Forum that was co-chaired by South Africa and the UK, we advocated for 2% of GDP to be ring-fenced for STI ecosystems. But given the magnitude of this issue, maybe more than 2% should be considered. Finally, through the implementation of Global Digital Compact and International AI Ecosystem should be established to provide open, reliable, and secure access to information and data. And the private sector should be— that investment is key, the investment for innovation and tech development for nascent AI ecosystem. I thank you.

Thank you so much. And thank you also for your leadership on G20 this year. And these same issues are being discussed there.

As well.

So, thank you. Next speaker is European Union.

Thank you, Chair. I have the honor to deliver this statement on behalf of the European Union and its member states. Let me first thank the President of ECOSOC for convening us today to discuss artificial intelligence and how it can help us achieve the SDGs. The global community must tackle AI challenges by strengthening scientific evidence, investing in talent and compute, building capacity, and establishing effective governance and regulation frameworks. Over the past month, the EU and its member states have been at the forefront of global efforts to make AI a global public good. Today, most developing countries face the urgent challenge of addressing AI divides while continuing efforts to close existing digital divides, a message that was underlined at the most recent IGF in Saudi Arabia. Key strategies to reduce AI risks and close AI divides include: setting ethical and rights-based governance, ensuring privacy, transparency, and accountability, involving diverse communities in AI design, investing in education and skills, and prioritizing linguistically and culturally diverse datasets. These efforts aim to ground AI in local needs, prevent the exacerbation of existing inequalities, and foster an equitable digital future— all core themes of the AI Action Summit co-hosted by France and India in Paris last February. To tackle these challenges and harness AI for sustainable development in line with the GDC, the EU fosters values-based, human-centric digital transformation through strong partnerships and multi-stakeholder engagement. For instance, we partnered with UNESCO to support the implementation of its AI ethics recommendation by helping member states build national institutions and developing a global ethics-focused AI toolkit for judicial operators. With ITU, we help countries shift from a focus on infrastructure to a holistic approach for meaningful connectivity. Together with ITU and UNDP, we help train officials and stakeholders to develop effective national digital transformation policies, offering global courses on emerging technologies, data governance, and AI. What the EU brings to the global efforts is also our own experience with developing AI regulation that fosters innovation. With the AI Act, the EU adopted a risk-based approach to protect human rights ensure user safety, and build trust in AI. It sets transparency rules for all general-purpose models, stricter obligations for powerful ones, and bans certain high-risk uses considered threats to safety and human rights. The AI Act's implementation supports building AI development capacity while boosting innovation and adoption by businesses of all sizes. It is with this background that we engage on the global stage, striving for international partnerships to support safe, secure, and trustworthy development and deployment of AI systems worldwide to help bridge digital divides and achieve the SDGs. Colleagues, in conclusion, the ongoing process co-facilitated by Costa Rica and Spain to establish an independent international scientific panel on AI and a global dialogue on AI governance provide a unique opportunity for the UN to step up its role in fostering AI partnerships. Through this process, the EU remains committed to ensuring AI is used to the benefit of all, leaving no one behind.

Thank you.

May I commend you for your timing? I had 2 seconds left. All right, the next speaker— thank you so much. The next speaker is Australia.

Thank you, Robert, and particular thanks to Rachel and Philip, and of course, thanks to the President of ECOSOC, Ambassador Rae. Australia recognizes AI's transformative power and immense potential to advance progress in achieving the sustainable Development Goals, and this will flow through to improve productivity, disaster risk reduction, healthcare, and climate change mitigation applications. I agree with Rachel that while AI has the capacity to deliver significant benefits, it also presents risks that may amplify harms and clearly may amplify disadvantage. From an Australian perspective, this would include bias, misinformation and disinformation, and increased opportunities for hostile interference. It's clear that persons and groups already in vulnerable situations may be disproportionately impacted by these harms. We must mitigate known and emerging risks to enable trust and adoption of AI, and we recognize how important this is for developing countries, where both the barriers and potential benefits are clearly amplified. We agree that effectively scaling AI innovations to contribute to achieving the SDGs requires equitable access, with a focus on the most vulnerable and least developed countries. To do this, and as Philip pointed out, we must support resilient digital infrastructure, build consensus around human rights, respecting guardrails that ensure responsible use, and address digital skills and literacy gaps. In this regard, Australia continues to work with Pacific Island countries to meet their aspirations for safe, secure, reliable, and affordable digital connectivity. And we do this obviously because these are the countries in our immediate region. These activities include investing in undersea telecommunications cable infrastructure, as well as support for cyber resilience, digital safety, and regulatory reforms in line with best practice. Global multi-stakeholder partnerships and capacity-building efforts are crucial for developing countries to leverage AI for sustainable development. We support multi-stakeholder collaboration to bridge the digital divide. We need researchers and innovators to apply their skills to ensure AI supports progress towards the SDGs. Balancing AI innovation and regulation necessitates close cooperation amongst policymakers, the private sector, academia, and civil society.

Thank you so much, Australia. Next speaker is the Holy See.

Thank you very much. My delegation welcomes this convening of this special meeting and would like to share some thoughts on how to ensure that ethical considerations remain at the core of the development and deployment of artificial intelligence. As we have seen, it is an extraordinary technological achievement that also poses many risks. For this reason, the international community has an important responsibility to regulate the use of it and in its many forms. To develop such regulation, it is essential to base it on common ethical principles that can also guide governance structures such as the Global Dialogue on Artificial Intelligence and the Independent International Scientific Panel. Indeed, A healthy policy on AI requires that technological innovations be placed within the largest projects of seeking the common good, and in this sense, the OLC considers it crucial to ensure that the development and use of AI always remain at the service of men and women, promoting fraternity and preserving critical thinking and the capacity for discernment. In this sense, the Rome Code for Ethics, issued in 2020 by the Pontifical Academy of Life, outlines a number of fundamental ethical principles such as transparency, inclusion, responsibility, impartiality, reliability, and privacy. This initiative has brought together many signatories, including governments, international organizations, other institutions, religious leaders, and private sector, and has been endorsed by many more. It is a concrete example of stakeholders coming together to define an ethical basis which will be essential for future regulatory frameworks, and to assist lawmakers in their efforts to balance AI innovation and regulation while seeking to promote the responsible and ethical use of science, technology, and innovations as agreed in the Pact for the Future, of which the Global Digital Compact is an annex.

Thank you.

Thank you so much. We have a lot of requests for the floor, but I want to pop back to the panel on just one issue, maybe for quick thoughts. One of the recurrent themes here is around capacity and skills. What is the— I mean, what are the kind of very key elements we need to all keep in mind? And, you know, as a delegate of the EU said, we are honored to work with EU support along with ITU on capacity— sorry, capacity building for policymakers and things like that. But what are we— what do we really need to do here? What are the 2 or 3 key things?

Thank you so much. And I'll start with— I think for me, it all starts and ends with leadership. Let me just stop there. And from the conversations you had, there's been some very deliberate, intentional, and very tactical sort of investments, whether it's through the EU Act or whether it's Australia. And thank you so much, Australia. We've been working on the Digital Earth Africa, for example, to look at geospatial and Earth observation. So very deliberate investments that are being made around capacity, around infrastructure, from undersea cable. And you know what's happening with undersea cables. Now, I think we'll have about 6 or 7 coming into our region. For me, I really want to go with that. There has to be deliberate leadership that is very focused on building the infrastructure for the industrial revolution. My last part, I think it's in the institutions. AI can come and go, but I think institutions need to be built so that institutions are resilient. The next innovation will be built on top of an existing institution. It won't be new. So, how do we modernize the institutions that we have?

Yeah, I will quickly jump in on that, Philip. I think building capacity, sustainable capacity within institutions of democracy in particular, will serve us well into the future. We run a series of courses on African AI ethics and African AI policy and human rights at the think tank I run in South Africa. And what's been effective about them is that they are led by African experts with African content, and that very often they're co-designed with those people that we're working with. So I think these are important principles.

Thank you both. All right, let's turn back to our speakers, and I can see we're not going to get through all the list right now, so just as fair warning, next speaker is Colombia.

Gracias, señor presidente.

Thank you, sir. Colombia is fully committed to the public regulation of artificial intelligence and emerging technologies, which is necessary if we're to manage the risks associated with its use. And if we're to seize its opportunities for the attainment of the Sustainable Development Goals and to protect human rights. The United Nations must play a key role in developing multilateral governance mechanisms that establish ethical and just parameters for the development and application of these technologies. And also to ensure that the digital gap that currently exists between and within countries is closed. It's for this reason that our country supports the creation of a scientific panel and a global dialogue for AI governance. The panel must be an inclusive environment. It must be meritocratic and multidisciplinary, where experts acting in their personal capacity but also representing a balance in areas of knowledge and regional contexts and periodically provide objective information to member states on the development of technology and on the potential risks and opportunities associated with its use. In turn, dialogue must be something that stimulates multi-stakeholder debate under the leadership of member states because we're the ones that have the responsibility to provide responses to the regulation challenges identified by the panel and to do this by way of the adoption of intergovernmentally negotiated policy measures. The panel must also stimulate the creation of mechanisms for capacity building and cooperation with developing countries. Nevertheless, even with this institutional setup, developing countries will not be able to enjoy the benefits of technology and of the digital economy in general if— and the speaker's microphone was cut off.

The time. Um, next speaker is India.

Thank you for giving me the floor. We thank the President for convening this important meeting. We also thank the speakers for their insightful briefings. India recognizes AI as a transformative tool for— to advance all 17 SDGs. Our India AI mission focuses on democratizing AI through the development of indigenous models enhancing computing capabilities, and supporting startups and research institutions. We have set up centers of excellence across key sectors such as healthcare, agriculture, education, and sustainable cities to develop AI-based solutions tailored to real-world challenges with clear alignment to SDG priorities. Guided by our Responsible AI for All framework, we are promoting the ethical and inclusive deployment of AI. Our digital public infrastructure covering over 1.4 billion people demonstrates how technology can be built on open, affordable, and scalable platforms. India is being regarded, regarded as a skill capital in technology and artificial intelligence with strong capabilities, talent, and policy frameworks. Stanford University has ranked India among the top 4 countries alongside the US, China, and the UK in its global and national AI vibrancy ranking. During our G20 presidency, India led efforts to build consensus on harnessing artificial intelligence in a safe, secure, and human-centric manner. The New Delhi Declaration— Leaders Declaration reaffirmed the importance of responsible AI and emphasized international cooperation to ensure its alignment with sustainable development. Our unique public-private models enables affordable access to computing resources for startups and researchers. India is ready to share its expertise to ensure AI drives inclusive and sustainable development for all. I thank you.

Thank you so much, India. Our next speaker is United Arab Emirates.

Thank you.

Thank you, Robert, and thank you to the speakers. I agree with the President of ECOSOC, Ambassador Ray, in that AI— there's no stopping AI.

And with its potential, with its huge potential for good, including in accelerating the SDGs, we certainly should be investing in this emerging technology.

But as was said by others, AI must be a tool for inclusion and not an engine that widens the global divides. And I'd like to offer 3 avenues for closing that potential gap. First is investing in education and building digital infrastructures. Both are foundational to capacity building, particularly in the Global South. Second, supporting South-South cooperation, cooperation between nations of the Global South not only builds capacity and fosters local innovation, but it also ensures that AI is developed and deployed with the specific needs and priorities for the Global South. And finally, and perhaps most crucially, bolstering multi-stakeholder partnerships. When it comes to AI, governments, academia, and the private sector can't operate in silos. We need these sectors to work together to ensure that everyone can participate participate in and benefit from this emerging technology.

Thank you.

Our next on the list is Croatia.

Thank you very much, and thank you very much for these interactive discussions. We don't see them very often, so thank you. So, we align ourselves with the EU statement, and here are some points in our national capacity. We are all aware of great benefits that AI brings to societies and to our everyday lives, and how important it is to achieve the SDGs. This is why it is important to prevent new as well as close the existing digital divides. Both technologies and development and ethical and normative framework need to go hand in hand. The security and human rights issues are arising from existing and new digital technologies and are truly multiple and intertwined. Well-regulated and inclusive AI is the one that benefits all. It needs to achieve gender balance and include notably older people, people living in rural areas, persons with disabilities, marginalized and vulnerable. It should also promote cultural and linguistic diversity. No one should be left behind. The United Nations has a pivotal role to play and we have the opportunity to remain the leader in AI governance. Basing our decisions and policies on science and evidence is the right path to AI that benefits all and is aligned with human rights and fairness. Digitalization is one of the 3 development priorities of Croatia. Croatia has been making progress in data protection and privacy, digital identity, the use of surveillance technologies, as well as online violence and harassment, and stands ready to share its experience including through its work in ECOSOC, if elected for the period 2026-2028? And lastly, if I may, a question to our panelists. What do you think we can do more as member states at the UN to work on bridging these digital divides and to make the AI trustworthy? Thank you very much.

Thank you, Croatia. We'll come back to the panelists in just a second with that question. We'll put it to them. We'll take a couple more speakers first, and next on the list is Mexico. Muchas gracias, señor.

Thank you very much, Mr. Moderator. Mexico recognizes that artificial intelligence, like all types of technology, is not neutral. It can widen existing gaps like we've already heard, or it can contribute to closing those gaps. For AI to really boost the attainment of the Sustainable Development Goals, we must ensure that its benefits reach all people. Inclusion will not be an automatic result of technological progress, but rather the result of intentional policy decisions. Scaling artificial intelligence in an equitable way requires firm commitments to international cooperation, accessible finance, and the generation of local capacities. In Mexico, we continue to encourage concrete initiatives such as public algorithm repositories, AI tools for SMEs, and programs that promote comprehensive digital literacy. For the Global South, it's essential to benefit from digital infrastructure, local research, and interoperable regulatory frameworks, avoiding AI concentrating power and wealth in just a few hands. Therefore, Mexico promotes global alliances. We participate actively in the Global Alliance for Artificial Intelligence, in the Group of Friends of the Recommendation on the Ethics of AI of UNESCO, and recently we hosted the regional consultations for the Americas on the Global Digital Compact. We are also promoting transparent, auditable, and explicable systems where people maintain control over their data. Ultimately, we need people-focused governance where we balance fundamental priorities, inclusivity and development based on the 2030 Agenda, on human rights, gender equality, environmental sustainability, accessibility, interoperability, transparency, innovation, multi-sectoral participation, and a future-based vision. Mexico is in favor of artificial intelligence that doesn't only transform our economies but also closes the gaps that divide us. and the Sustainable Development Goals that seek to enhance international cooperation and ensure that the benefits of technology in general reach all people. Thank you very much.

Thank you so much, Mexico. And we'll take one more speaker and then we're going to come back to the panel for some reflections, including the question from Croatia. Next speaker is El Salvador.

Thank you very much, Moderator. Distinguished panelists, in a country that has placed the digital transformation at the heart of its development strategy, El Salvador believes that the opportunities that stem from the uses and application of artificial intelligence are things that cannot— must not only be capitalized by just a few global stakeholders. AI must be turned into a concrete tool to resolve public problems, to strengthen and boost productivity, and to boost opportunities for our societies. Scaling the positive impact of AI requires more than just access to technology, cutting-edge computation, data, and other elements of the digital ecosystem. It requires adequate regulatory frameworks, local talent, international operations focused on results and political will in order to act with a vision of the future. El Salvador has made significant steps forwards to lay the foundation of a robust digital economy. The recent law on artificial intelligence and technology promotes their responsible development and encourages investment development and stimulates training up talent, in particular through public-private partnerships. Building our technological actors. We are also making progress in designing governance mechanisms that enable us to guide artificial intelligence governments in an ethical and inclusive way focused on public entities. We believe that no country can make progress alone, and we need global policies that respect national priorities and that encourage the public use of the right technology. And this is the true way of democratizing the results of artificial intelligence. Thank you very much.

Thank you, El Salvador. Okay, panel, you have a magic wand. Croatia said, what can we as member states do? What would you have member states do?

Well, thank you very much for, for such an important question. As I think I indicated earlier, I think the scale of international cooperation needed far outweighs anything that we have dreamt of before. So, I think it's incumbent upon all of us, whatever sector we're in, whatever stakeholder we represent, to think carefully about how we can contribute towards a shared agenda around closing the digital and AI divide. One of the efforts that I think have been really effective is the effort being led by UNESCO to support the implementation of AI ethics and the recommendation on the ethics of AI. And they did this through localized, multi-stakeholder assessments, roadmapping in countries, which brought together a range of different actors to learn from one another and to build together a shared vision and trajectory and policy path for action.

Thanks so much. I think I'll be very quick. I think for member states, that's why I wanted Rachel to start, because I'm part of a member state. I think one is, again, and I'll say this again, it's about leadership. And His Excellency the President of Kenya has been very clear that AI is a whole-of-society approach. It can't be— we have to break the silos because the burden of history, and recognize this, of the Global South has been exclusion from opportunity. And if AI is a general-purpose technology, we cannot look at it like previous technological ages. So how do we— must put all our resources so that everybody has access and it can transform I think the second part of your question, and it was how can we make AI more trustworthy? And I think for me, it's these conversations that we need to have, and we need to work together in the way that we deliver the promise of AI, which is about solving health, solving climate, solving food systems, solving education. Because if people do not see the benefits of AI, then why would we trust it when we cannot see its benefit translated to improving our lives and SDGs? Thank you.

Thank you very much, both of you. Okay, we'll go back to our speaker list. We're starting to run out of time. The speaker list, I understand, will continue into the next session, but we'll take a few more before I have to close this session and we move on. Next speaker on the list is Dominican Republic.

Thank you for giving me the floor, and thank you for your valuable comments and for organizing this special meeting.

Transform sectors like health, education, environmental sustainability is huge. However, in order for these innovations to be implemented at a large scale, we must build the local capacities we need and ensure equitable access. This requires solid global cooperation where developed and developing countries work together in order to overcome the barriers of infrastructure, skills, and access to data. To data. This is the only way we'll be able to ensure that AI-based solutions are not limited to certain sectors or regions, but rather that instead they reach the 4 corners of the world. It's essential that in this process we don't only promote innovation, but we also protect fundamental values such as ethics that will guide their implementation, and also inclusivity and transparency. It's here where global partners including governments, the private sector, academia, and civil society must strike the right balance between driving this technology and innovation by establishing regulatory frameworks that respect human rights and that promote responsible AI. In the Dominican Republic, we understand that this balance is not an easy one to strike, and that's why we're in favor of public policies that encourage accessible, inclusive, and transparent AI, ensuring that the most vulnerable communities are not left behind. Thank you.

Thank you so much, Dominican Republic. And we'll go straight on to, uh, international organization ILO.

Hi, thank you, Excellency and colleagues. Um, for the ILO, the key question is not whether AI is transforming the world of work, but how we shape that transformation to uphold decent work and social justice. ILO findings show that while generative AI may augment more jobs than it automates, this promise is deeply unequal. As mentioned by the panelists and by others, in many developing countries, weak digital infrastructure, limited access to training, and constrained institutional capacities continue to hinder the ability of workers and enterprises to fully leverage AI for productivity and inclusive growth. And women, especially in clerical and routine roles, are disproportionately at risk because they fill the jobs most at risk of automation. Governance is key. Through the ILO Global Observatory on AI in the World of Work, we are generating evidence and policy tools to guide inclusive responses, focusing on algorithmic accountability, fair data governance, and social dialogue. The observatory underscores that labor markets will only benefit from AI when workers are empowered to shape its deployment. Today's digital divide could become tomorrow's structural fault line. That is why the ILO urges investment in 3 enablers: digital public infrastructure and interoperability to ensure developing countries can participate in the AI value chains; human capital, with a particular focus on reskilling women, youth, and informal workers; and institutionalized social dialogue to embed ethics and equity into workplace AI use.

We also—

the I echo the call of the Global Digital Compact.

AI development must not be steered solely by commercial incentives. Public-private partnerships, community-driven innovation, and intergovernmental coordination are essential to ensure AI advances the Sustainable Development Goals. Thank you.

Thank you so much, ILO. Next is Liechtenstein.

Thank you, Moderator.

As we have heard from many speakers this afternoon, artificial intelligence is rapidly transforming our world. AI technologies offer tremendous benefits for scientific progress and sustainable development, but they also risk being abused. The fast-paced AI advancements must be addressed proactively, also by ensuring respect for human rights and fundamental freedoms at all times, as well as that they benefit all of humanity. Liechtenstein therefore calls for closing the digital divide within and between countries, including the gender digital divide. In this regard, we welcome the adoption of the Global Digital Compact last year as a key contribution towards AI governance that also fosters an inclusive, open, safe, and secure digital space that respects, protects, and promotes human rights. There is an urgent need to translate the GDC into action and to further strengthen cooperation in the field of digital technologies and AI governance. During Liechtenstein's presidency of the Council of Europe, the Framework Convention on AI, the first-ever internationally legally binding treaty on AI, was adopted. The convention aims to ensure that AI systems respect fundamental human rights, uphold democratic principles, and adhere to the rule of law.

It can serve as a positive example for AI governance that ensures ethical standards while aiming to promote innovation.

Unfortunately, we witness increasing malicious use of AI, for instance in the form of AI-powered cyberattacks.

With wide-ranging effects for international peace and security. This underscores the imperative for clarity regarding the application of international law to malicious cyber activities. It is clear that cyber warfare, and including AI-powered cyber threats, is subject to international law, including international humanitarian law and international criminal law, inter alia under the UN Charter, the Geneva Conventions, and the Rome Statute of the International Criminal Court. In the area of However, in the era of rapid technological advancements, it is imperative that the digital space remains a tool for progress and human dignity rather than harm and division.

I thank you.

Thank you so much. And we're really starting to run out of time, but we'll take, I think, maybe, if we're lucky, 2 more speakers. Next one is Georgia.

Mr. Moderator, all protocols observed. As the ICTs have impact on our lives, the development of the digital economy is one of the main priorities of Georgia. Enshrining the right to access to the internet in its constitution. The emerging technologies and the AI plays crucial role in the process of building the digital economy and information society at the local and global level. The AI's escalating influence extends across various sectors, which is believed to bring further significant changes in our everyday lives. Governments, businesses, and society must quickly adapt to the new layouts of economy. But if missed, the best opportunities will be captured by frontrunners. This will further deepen the current digital divide and exacerbate the shortages of skilled workers. Stemming from above mentioned, in 2024, the government of Georgia designated AI as one of its top priorities in the field of technology and innovation, which later translated into the strategic priority of Georgia identified under the Digital Economy and Information Society Strategy 2025-2030. This strategy sets the promotion of trustworthy and secure AI services as one of Georgia's key priorities. This strategy also emphasizes the importance of addressing emerging challenges during the creation of digital policies. In response to these challenges and in line with the recent developments such as the European Union's AI Act and the Council of Europe's Convention on AI, Human Rights, Democracy, and the Rule of Law, the Georgian government is actively working in close cooperation with the World Bank to develop Georgia's national AI strategy. It will lay the foundation to the country's first regulatory framework for AI, addressing associated risks and challenges while also supporting the advancement and potential of Georgian AI technologies. To ensure the implementation of these objectives, the strategy's action plan outlines specific actions including the preparation and adoption of an AI development roadmap the establishment of an AI research center, and the development of Georgia's national AI strategy. Preparatory work is underway to establish a national AI research center, which will aim, among other objectives, to develop and implement Georgian-language AI solutions in the financial, healthcare, academic, and agriculture sectors and support various research initiatives within the AI industry. Strategic vision and implementation plans are crucial. Our national AI strategy will incorporate input in how we drive economic development and improve the well-being of its citizens.

Thank you.

Excellent timing, Georgia. Thank you so much. We'll take one more speaker from the floor. I'll turn to the panel for a quick comment, and we'll close this session and move on to the next one.

France.

Merci beaucoup.

Thank you very much, Robert, for giving me the floor. As was said by the Special Envoy of Kenya, artificial intelligence is not a new thing as such. What is new is how it erupted on the international seen. And this makes it possible for us to make into a global public good. And that's a priority for us in France. The benefits are gigantic, especially for the developing countries. The issue is the accessibility for the major part of the world to these things in a world where it's developed in a— by a small number of major companies, a small number of countries. So to transfer AI into global public good, we organized the 3rd sequence of the AI Summit action in February last. It was co-chaired by us and India in Paris, and we brought together a large number of stakeholders to define the major principles for the partnerships we need to make sure that artificial intelligence is ethical, inclusive, transparent, safe, secure, and trustworthy. During this summit, a public-private foundation was created which is called Current AI. It's just started, and its goal is to support the financing of trustworthy infrastructure, allowing access to reliable databases, and to develop AI for climate, healthcare, and education. For this, we need global governance— global governance which is rooted in a scientific consensus, a common approach. And that is why France supports the idea that were developed in the Global Digital Compact to establish an independent scientific panel and dialogue on global governance for artificial intelligence. Thank you very much for giving me the floor.

Thank you, Franz. And we are short on time. You can see that we're all looking at the clock here.

Okay.

So, I'm going to wrap it up with a question to the panel. And, you know, again, one of the themes that has come up on a number of occasions has been around the role of the private sector. And I I see Chris from Microsoft here, who I noticed as I was looking in this direction. But we have many other levels of private sector. There's global private sector, there's local private sector, and obviously, the technology, by and large, comes from private sector. So, what are the most effective ways of collaborating with the private sector as we talk about trusted, safe deployment of artificial intelligence? intelligence. Thoughts from the panel?

Philip?

Very quick, because I work with Microsoft Beta and a lot of these folks. I think the first thing is collaboration. It's— you really have to work together. And what we've seen, for example, for what we call growth markets is that this is also new for the private sector. And so, how do you ensure that as you build, then you lay the guardrails? Because part of the trust issue has been that governments have always tried to catch up and private sector is always innovating and running ahead of innovation. So, I think for us, we've seen in our experience, it has to be collaborative, and you build the guardrails as you go forward. But the second thing is also this proactive disclosure. I mean, part of trust is being able to disclose the roadmap of where the technology is going, but also including local developers who are on the continent to help build these products so that they're more nuanced and respond to the challenges. I think finally, finally, finally, I think it'll be important to be comfortable with uncertainty. Because private sector is very good at developing in fast cycles. So, how do we ensure that public sector also upskills and develops at the same pace?

Thank you so much. I think it's important to recognize that practice around responsible, ethical, and inclusive AI starts with the private sector, which is not a homogeneous group. And there's huge divides within what we're talking about with the private sector. When we're talking about big tech, I think it's important to request access to the data that they have. I think they hold huge amounts of data that's very, very valuable for local innovators. And I think there's an importance around open-source protocols to build these more inclusive and accessible AI ecosystems.

Thanks. Thank you so much. And with that, I'm going to close this session off. I do want to make one little plug for our own organization, UNDP, which today launched the Human Development Report, which we produce every year, but this year is on the subject of AI. And it was launched by our administrator in Brussels today and is a global launch and available on the UNDP website. So, the focus of that report is really about the choices as humanity that we need to make, the intersection here with human development, more from the demand side than the supply side that we often focus on. So I encourage you all to look at that. I want to thank my distinguished partners here, as in their incredible insightful comments. As always, it's such a pleasure. And thank you to President of ECOSOC, Ambassador Ray, and the Mission of Canada for having us here as well. And with that, we'll close this session. Thank you so much.

Every crisis brings urgency. Every second matters. Every dollar counts. That's why the World Food Programme is leveraging artificial intelligence to save lives smarter, faster, and more effectively. With artificial intelligence, we can predict food shortages 60 days in advance, allowing WFP to mobilize assistance before hunger takes hold. When disaster strikes, AI-powered tools can help WFP analyze drone imagery in hours, not weeks, so relief reaches communities sooner. And in conflict zones, AI helps ensure food reaches the right people safely, efficiently, and with greater accountability. But as AI's role in humanitarian response grows, so do the risks and responsibilities. That's why the World Food Programme is launching its first-ever global artificial intelligence strategy, a pivotal step in ensuring that we respond quickly and responsibly. With the right guardrails, ethical safeguards, and partnerships with tech leaders, we are setting new standards for AI that puts people's needs at the forefront of every response. By 2027, AI will be It will be integral to WFP's operations, helping us act faster and more effectively in crisis zones to alleviate hunger and save lives. This is artificial intelligence for a world without hunger.

Digitalization and AI have brought groundbreaking innovations to help protect the health and lives of workers. Today, robots are operating in hazardous environments, doing the heavy lifting, managing toxic materials, and working in extreme temperatures. They take on repetitive and monotonous tasks, while digital devices and sensors can detect hazards early on. However, digital technologies can lead to accidents, ergonomic risks, work intensification, reduced job control and blurred boundaries. Not forgetting the safety and health risks for workers across the digital supply chain, including those in extraction processes to those powering AI and workers in electronic waste management. Nevertheless, we believe that through human-centered policies, we can ensure that digital innovations are drivers for safe and healthy working environments. environments for all. To learn more, check out ILO's new report on revolutionizing health and safety through AI and digitalization.

Hello, good afternoon, and welcome to the second part of this exciting ECOSOC session on AI. And a really big thanks to Ambassador Bob Rae, the President of ECOSOC, for continuing this tradition of having this discussion on artificial intelligence in these ECOSOC chambers. And indeed, you know, my colleague Rob Opp had led a very insightful session of how the emerging technologies and AI, you know, are transforming our world and what they can bring to us. I think in this now session, we'll explore how to bring the opportunities of them to everyone so that it's not contained to a few countries, a few locations, a few geographies, a few companies. And I have I have an exciting panel of colleagues here to contribute to that. So, I'm Thomas Lamanouskas, Deputy Secretary-General of International Telecommunication Union. I would like to say UN Agency for Digital Technologies. And again, this panel— I'm joined today here by a number of speakers representing the multi-stakeholder environment that really is representative of the AI ecosystem. So, first, I'll give you 1 minute, I was told by my script, for each of you to introduce yourselves. So, you know, we'll start with Jerry John Konyo, the Head of Responsible AI Lab, Kwame Nkrumah University of Science and Technology in Ghana. So, Jerry, please introduce yourself.

Thank you. My name is Professor Jerry John Konyo from the Kwame Nkrumah University of Science and Technology, Ghana. I lead the Responsible Artificial Intelligence Lab which is one of the AI4D labs led by IDRC and FCDO. I'm also the project lead for AI for Sustainable Development, and I'm glad to be here. Thank you.

Thank you very much, Professor Okonjo. Now, I also have Ms. Omaima Makhlouk, Data Focal Point for the Major Group for Children and Youth. Omaima, please introduce yourself.

Thank you so much, Tomas. Hi, everyone. I'm Omaima. I'm Moroccan and French, and the Data Focal Point for the Major Group for Children and Youth. I work in tech as a data scientist, as well as a data consultant with the World Bank, and I'm truly excited to be here with such an exciting panel. Thank you so much for having us.

Thank you very much, Jemimah. And last but definitely not least is Chris Sharrack, Vice President for UN and International Organizations of Microsoft. So, Chris, over to you.

Thanks, Thomas. It's a pleasure to be here. I'm Chris Sharrack. I run the team in Microsoft that engages the company with the UN and with international organizations around the world. including, for example, I chair the business at the OECD Digital Policy Committee.

So, thank you. Thank you very much, Chris, for that. So, indeed, just to kickstart this, just a few comments from my own side. So, indeed, when we look about development of AI, implementation of AI around the world, we see that actually, in a number of emerging markets, the optimism about artificial intelligence and emerging technologies is even higher than in the developed world. Actually, by some estimates, you can see that the people who are more positive about AI rather than negative in emerging markets are double compared to Europe and, let's say, the developed world. So, there is a lot of opportunity there. At the same time, AI also, when you look from the negative impact side, AI sometimes impacts less. Now, our colleagues here of ILO, who've been featured a lot in this session as well, their studies show that if, on average, You know, around 5.5% of the jobs might be, let's say, automated out. In the developing world, only 0.4% might be automated out. So I think the risk is less so, and of course this is also because of the nature of the jobs. And clearly the opportunities are there as well. However, to make these opportunities a reality, we need to overcome a number of barriers that we'll discuss here today in the panel, and we'll see how to do that. We need to overcome the barrier of infrastructure, and again, you know, like very often quoted figure that 2.6 billion people around the world are still offline, regretfully. And a lot of them are actually concentrated in the low-income countries. Actually, in low-income countries, less than a third of the population is only still connected. Of course, when we talk about infrastructure, we also talk about the energy infrastructure that is crucial for the digital infrastructure to function. And when we talk, we also talk not only about connectivity infrastructure, but data centers and other facilities, which is still very unequally distributed. Just for comparison, in Africa, which today has nearly 19% of the world's population, there's only less than 2%, actually 1.8% of global data centers are there. So, it's definitely a huge divide. We also talk about skills divide, and in that way, also the divide where in the value chain most of the people are contributing to, whether they are PhD holders and data scientists or they're data labelers. And I think here also results in a huge economic divide as well. We also have a huge policy divide, with still over half of the countries not having AI policy, and actually only 15% of the countries that have AI regulation. And this policy divide also impacts how much we can develop AI infrastructure and ecosystem in these countries. We have these inequalities on a global level and between the regions, but we also have these inequalities within the regions, for example, including in Africa, where a few centers in Africa hold most of the data centers where they're really concentrated. So, it's not just about resolving those inequalities, let's say, between Global South and, if you will, Global North, but also within Global South as well. So, how do we do that? That's what we'll try to explore today with our panelists. And we will start, actually, Jerry, with you, if you don't mind. So, you know, Jerry, you are leading several initiatives on responsible AI and digital development across Africa. So, from your experience, what are the most critical elements that are needed to build a sustainable AI capacity? And how can international and regional partnerships be Health Minister Ngot. Please, over to you, Jerry.

Okay.

Thank you very much, Thomas, and thank you to all of you. Let me begin by drawing our attention to the reports by the UN High-Level Advisory with reference to what needs to be done as far as AI is concerned. Certain key enablers were pointed out, and they've been mentioned severally as far as today's interactions are concerned. Key enablers that we need to pay attention to, beginning with infrastructure— and by infrastructure we are talking about reliable internet and reliable power. Irrespective of how good your AI solution is, if there isn't an infrastructure to deploy it on, there will be a problem. What we have realized is that it is important that we don't limit ourselves only to the urban centers. A significant percentage of the population in the Global South find themselves in the rural communities. What that means is that we must be intentional in ensuring that, irrespective of people's location, they have good quality infrastructure with which to roll out AI solutions. The second issue borders on the availability of data. Your solutions in the AI space are as good as the data with which your models are trained. It's important that your data is of good quality it is important that your data is representative. If your data does not cut across, what it means is that your solutions are going to heighten existing stereotypes, and that is the reason why particular attention must be paid on that. There has been a lot of conversation with reference to compute, and I wouldn't want to belabor the point, but it is also important that we pay attention to talent, and by talent, I am not just interested in developing. It is important that the end users are skilled as far as the use of AI solutions are concerned. We can make a lot of impact in the public sector if we pay particular attention in building capacity in that regard. Having said this, now we come home to what are some of the examples of interventions that we have seen that we can ride on. Rachel did mention the AI for Development initiative within Africa and other parts of the Global South. What we realize is that we are making giant strides as far as growing AI in the policy space, in the innovation space, and in the research space is concerned. As a result of some of these initiatives also, a number of countries are actually developing their AI strategies. Some of them have already published their strategies. Ghana, we are in the process, and come June, hopefully, we will put our strategy out. Now, the point that I would like to make is that all these things that we have talked about— infrastructure, data, compute, and talent— must be tied into a strategy. It can be a continental strategy, it can be a national strategy, but a strategy alone does not make a lot of difference. It is important that we pay as particular attention to the implementation of the strategy as we pay to the drafting of the strategy. And by way of this, what I mean is that it is important that we make commitments as far as budgetary allocation is concerned with reference to the driving of this strategy. If we really want to make a lot of impact in the areas that I've talked about, it is important that year in, year out, we make allocation with reference to budget when it comes to infrastructure, when it comes to data, when it comes to compute, and when it comes comes to talent. The last thing that I will say is that it is good to have development partners' support as far as some of these things are concerned. The reality is that it is not sustainable and we need to demonstrate commitment by member countries making budgetary allocation to drive innovations and initiatives like this.

Thank you. Thank you very much, Jerry, for such a clear explanation of the framework, including elements of infrastructure, talent, compute, But also, importantly, strategy. And as I say, the strategy is only the strategy if it's backed up by resources as well. But also, back to the talent piece, and I think that's where I'll turn to Maima, because of course we hope that this talent is coming up from the young people from across the world, actually, but the Global South actually has proportionally higher youth populations. So, it would be great to hear that perspective, and especially drawing upon the discussions that happened in ECOSOC Youth Forum among young leaders and experts on Global Digital Compact, on Declaration on Future Generations, and how to achieve them? And what are the insights from the young leaders in terms of how can we shape in a positive way AI development?

Over. Yeah, that's a really good transition. I think, you know, from the ECOSOC Youth Forum, something that came as being crystal clear from young leaders is the fact that we need to shape and co-create the future of the digital agenda, right? It's not just about shaping it around us because we're the next generation. Something that I want to mention, especially with regards to Global South, you know, coming from Morocco, is that it's not just about inclusion and it's not about justice. It's just about empowering young leaders in Morocco and other countries to be innovators, creators, and not just be just consumers of technology. You know, just like you mentioned, more than 2 billion people are offline and actually more than only 40% of youth people, young people in the Global South actually have access to the internet. So that's like really an issue, right? But despite that, we have excellent innovators who are contributing in that space. For example, during the Youth Forum, we heard from Marina Al-Khawand, who back in 2020, after the Beirut explosion, created MedoNations, a platform that optimizes the medical resources allocation between donors and patients in need, leveraging AI and data analytics. That shows truly the talent of youth. It's just really about infrastructure. It's about digital literacy. It's about all of these important things that we need to give youth to be able to create more and to be included more in this sense. Because, you know, if we fail, AI is just going to deepen this inequality and digital divide even more instead of like helping us solve problems, right? So that's another thing to consider. Now, after the Youth Forum, I think that 3 main opportunities emerged, at least from our discussions. The first one is that, as I mentioned, youth are already leveraging AI to solve issues in their countries, in their communities. But what's missing? You know, it's not ambition or talent. It's truly, as I mentioned earlier, it's funding, it's scaling, it's mentorship. And that's why we are all here today, right? It's to answer questions like Croatia asked earlier, like, what can member states do? And that's exactly what you can do. It's to reach out to your local youth communities that want to innovate and help give them what they need to do that, right? To be able to scale their innovations and solutions. The second thing is that, you know, we need to call for an education reform. As I mentioned earlier, there are so many things happening in the AI space, and we need to make sure that education follows. So we need to include coding and ethical tech into the curriculum all around the world. to be able to kind of ramp up. And lastly, something important, and I'm happy to be here today in this panel, is that youth must be included, right? I know there are a lot of meetings like this one, or like, for example, the UN Secretary-General's Advisory on AI Board. So hopefully we can be here and we can give our perspective, however advanced we're, you know, we're not maybe experts, we don't have years of experience, but we still have a positive experience to give because participation without power is tokenism, and we're done with that, really. So yeah, to break this cycle, please, Please, I'm calling open the member states here present to kind of help us, work with us so that our work can be resource-skilled and embedded in real policymaking. Lastly, I just want to say that we need to shift the mindsets. We always talk about AI for the South. We need to shift it to AI with the South, for the South, and global. It's not just about Silicon Valley and some of the major tech hubs or tech companies. It's really about literally anyone who has some interest in AI and technology to use it for good. So, really hope that we can shift mindsets and include everyone in the process. Thank you.

Thank you very much, Ma'am. And I really like your last point, you know, because I think it definitely cannot be about producer community on the one hand and user community on the other. It's about inclusive development for ourselves, you know, where we have our own skills and the capabilities to deliver, to find our own solutions with the tools that we have here. And I agree, there is a great potential with the youth, but they need to be empowered with infrastructure, with the skills, skills and education and with the inclusion. I think part of that empowerment investment is, of course, private sector, because we cannot expect all the development coming from the governments, especially— and tech sector is the sector where we have seen the success of the private sector-led development around the world. Chris, from your perspective, what is the role of the private sector, from your perspective, in ensuring this inclusive development of artificial intelligence, emerging technologies, around the world, including Global South? And what is the role of multi-stakeholder partnerships to achieve that?

Well, thanks very much, Thomas. It's a great question. And I have to say, it's really a privilege to be here. It's always a pleasure to take part in the STI Forum, partly because it is a multi-stakeholder event. And as I was walking down the corridor, to come in here, I was thinking about the last time I took part in this, which was 2 years ago, and I recognized the South African Ambassador who referred to it earlier, the 8th STI Forum, which took place in 2023. And it was a couple of months after LDC5, where Microsoft had the privilege to co-chair the Private Sector Forum, and it was a few months after the launch of ChatGPT and Copilot for Microsoft and the other different forms of generative AI. In some ways, the world was quite different, but in many ways, the conversation was very, very similar. The conversation then was about the role of digital technology in driving the SDGs. It's now about the role of AI in driving the SDGs. And I think, in a sense, what's changed is that the stakes are higher, precisely because we see, you know, so starkly with generative AI the potential positive impact. I think everyone uses this figure of 80% of the SDGs that could be accelerated. I'm sure it's 100%. And at the same time, the UN has also risen to the challenge, to a large extent, through the Global Digital Compact. In actually addressing the policy questions around this technology. So, you know, here we are in 2025, still facing the same question, but maybe with a little bit more resting on us all. And I think that a general-purpose technology like this does not come along every year. We have to think about this in the same way that we think about electricity. This is a technology that has the potential to change every aspect of economic activity, every industry, every sector. And if we don't harness it and make sure that it is diffused and adopted across the world, then we risk finding ourselves in the same situation we do today with electricity, where still 700 million people don't have access to electricity. So, you know, how do we address that? Look, the private sector has has got a role to play, and I can speak to Microsoft's activity. We've been present in Africa for 30 years. We've got a presence in 12 countries. We work through a network of partners. We drive out our investments as far and as fast as we can. Clearly, we're in it, we're in business, so we're looking for sustainable activity that will deliver a profit, but we're making investments. Now, look, the scale of the investment cannot be underestimated. A handful of American tech companies are investing hundreds of billions of dollars in investing in building data centers and other infrastructure. Microsoft, we've talked about the figure of $80 billion. We invested $1.5 billion— we have announced the plan to invest $1.5 billion in Kenya. And I'll use that as a bit of an example to build the data center there. So, you know, we are making the strides where we can. But clearly, it takes a multi-stakeholder approach to bridge the gaps that exist. We can think in terms of a 3-legged stool: government, the private sector, and civil society. I don't talk about public-private partnerships. I talk about multi-stakeholderism. All have a role to play in this regard. So, 3 areas, maybe, just to touch on. First of all, connectivity, and I have to recognize the Partners to Connect program that ITU has led that has really been so transformational in actually unlocking new input and new commitments from stakeholders from across the sectors. If we don't connect the world, then this technology will remain out of reach, and that has to be something that we come together to deliver. The question of investment, of course, the private sector needs to mobilize capital and invest, But at the same time, the public sector has a responsibility to create a business environment that will be conducive to investment. Now, I know that's a familiar message, and I'm sure this will be something that's picked up at Financing for Development 4 later on this year. What does that look like? How can things like blended finance maybe help de-risk investments? We have to think about those questions with the lens of digital, infrastructure. But one aspect, to come back to the Kenya example I cited earlier, is that the demand has to be there to justify the investment, and there are things that the public sector can do to help that. One is through adoption. If the public sector is an early adopter of the technology, then that helps drive the demand that then makes the investment more viable for the private sector. But there's also the question of aggregating demand across regions, and I'd like to just maybe, before closing, refer to some work that the East Africa Community are doing as maybe something we could draw a little bit of inspiration from. I don't know off the top of my head the population of Kenya. I'm sure the Kenyan delegation can tell me. I think it's about 40 million. If we're going to justify an investment in that country, It will be all the more viable if we can also have the demand for usage of that data center from countries surrounding Kenya. And the East Africa Community is actually tackling that question now. How could a trusted regional data zone be created that would allow cross-border data flows to be done with trust and security that would then justify an investment like the one that we're that we're planning in Kenya, for example. And I do hope that these conversations on data governance, which Gerry referred to— data is so important. And the CSTD in Geneva just last week kicked off this process that was triggered by the GDC. I hope that these questions of data and data governance can come more to the fore, because really getting a framework around data governance will do so much to give confidence to the private sector to invest and for that demand to then come through. I'll stop at that point.

Thank you very much, Chris. Indeed, we need the private sector to invest, but we cannot push them to invest. We actually need to look at that as a partnership way. Actually, we, together with the development finance institutions, are trying to achieve that through our Digital Infrastructure Investment Initiative, exactly that way, how to make sure that there is enabling— the countries know what is enabling environment for to attract the private sector, and of course, in that time, then we would hope and expect the private sector will step up on their part of the equation. Continuing the conversation, with a few more minutes left here, and coming back to what Omayyama said about this inclusive local development, so that we don't just have something developed from some centres of the world and then they're just consuming, so Jerry, I will come to you now in terms of with your experience as a leading voice on responsible AI in the region. How do we ensure that AI is developed everywhere, including in Africa, for Africa, and in a responsible, locally relevant way? What are the things we need to do to achieve that?

Thank you very much, Thomas, and thank you to my co-panelists for such insightful deliveries. Responsible AI has tremendous potential to change the lives of many people globally, and it's important that we pay particular attention to that. And remember, I keep stressing on responsible AI and not just AI in this generic context. Responsible AI is people-centered AI. Responsible AI focuses on adding value to the quality of life life of a person and adding value to communities. And so, uh, in rolling out responsible AI solutions, first and foremost, it is important that we pay particular attention to the intended users and bring them to the table. And that is the more reason why stakeholder engagement is crucial in ensuring that AI solutions that are deployed in many areas are relevant to the community. If you don't have the people at the table at the design stage, at the end of the day, we will develop solutions that will not be helpful to the people for whom it is intended. And so it's always the philosophy of the lab that I lead to ensure that in every development that we come up with, we have the stakeholders on board. And so where you are developing solutions for farm commerce, or you are developing solutions for people in the health space, you always want to hear the voices of the intended users. And so stakeholder engagement, very, very important. The other thing has to do with ensuring that we are responsible in the curation and the management of our datasets. I've already indicated in my earlier submission that data is what runs AI, but the fact that data is so crucial does not mean that we should use irresponsible means in curating the data and managing it. For instance, in applying AI in the health space, it's sensitive human information that you are working with, and it's important that we are very responsible in the use of them and seeking various forms of consent as far as the person that we are working on is concerned. The other thing that I want to talk about with reference to making sure that AI is inclusive and relevant and adding value to the quality of life of people is ensuring ensuring that in rolling out AI solutions, we don't create a class system. It is important that we develop solutions that are affordable to all manner of persons. If we develop AI solutions and only the rich can afford, we will have a problem on our hands. And so we must be conscious of cost as far as AI solutions are concerned. 1.2 billion of the world's population are persons with disabilities. That is 15% of the global population. And in many conversations around AI, we actually do not bring such people to the table. Responsible AI requires that we ensure that when we deploy an AI solution, we don't disadvantage people further. And that is the more reason why we must be conscious of ensuring that we secure the views of persons with disability and ensure that we are deploying technology to bridge the gap rather than widen it. The last thing I will say, Thomas, and then I'll hand over to you, is that frameworks are important. Globally, several frameworks have been rolled out, but who is using them? It is important that we develop and deploy avenues through which intentionally people can use these frameworks. Within the AI4D ecosystem of labs, we've developed what we call the fastest framework framework, facets for fairness, accountability, confidentiality, ethics, and transparency. And what we do is that any solution that we are developing within the ecosystem, we use the framework as a lens. And how it helps us is looking at the responsible AI principles where some scores are low, it gives us an indication that we need to pay particular attention as far as those principles are concerned. In my thinking, we should be able to make giant strides in responsible AI if we avert our minds to these things that I've talked about. Thank you.

Thank you very much, Jerry. And you mentioned how to make AI inclusive for everyone. And I think, Omayma, from your perspective, how we make it— how you would see we should make it inclusive, including for different communities and groups, including women, youth, marginalized communities, and underserved regions, please.

Yeah, well, I'd like to say first that actually AI is not inclusive by design. It's only inclusive if we design it to be intentionally. And that's really important, right? And I totally agree with everything you mentioned with regards to especially like the cost of it in particular. I know we discussed this also during the last panel. AI is compute heavy, it's expensive, and sometimes not even accessible to the people that need it the most. So that's why, you know, it makes the digital divide even like stronger. And sometimes actually AI shouldn't be used at all, right? It's all a question about like who should be using it for what purpose and for what reasons, right? Because something that is important to mention is is that we need to ask ourselves, like, how we collect the data, how we evaluate the models, and who gets to define what success criteria means when it comes down to leveraging AI for a specific issue. And I think the best way to kind of make this process inclusive is by co-partnerships and, you know, co-working. And by that I mean, for example, at the bank, usually when we work on specific use cases, especially, I know, like with my work with the poverty equity practice, we tend to involve local talent, we tend to local resources because the goal here is to make sure that we have frugal models, lightweight, that can run under any circumstances and any conditions because they reflect real-world context. Secondly, with regards to empowering youth and marginalized communities, we need mentorship. We need funding. We need education. For example, in Morocco, we have the UM6P Explorer Program in collaboration with MIT Sandbox that connects students with older, more expert AI engineers to help them build potential solutions. The same goes at Stanford with the AI for Good Institute where I'm involved, where we also try to kind of teach some of those core concepts of AI to lead to rapid prototyping. So I would say like it's really time to focus on kind of treating youth-led innovation as research and development and investment in the future, not just like charity, I would say. Just like we invest in biotech, we should also invest in AI for good. Thank you.

So thank you, Rensha Maima. And then to close this round of comments, Chris, from your perspective, maybe from Microsoft's perspective, what are you as a company doing to actually live up to those words that you said, you know, like to really develop, deliver for Global South and Africa and other regions as well?

Yeah, well, look, I talked about the Kenya investment as an example of where we're, where we're partnering with G42 in that, in that, in that case, looking to build something that will alongside it have a skilling program that will draw on renewable energy to power it and will also, which address some of the language questions that are so rightly at the front of people's minds in terms of training the models on local languages like Swahili. So there's that. And then, as I mentioned earlier, the connectivity work we do with Airband, where we've had commitments for a long time now to deliver connectivity in Africa. And of course, on the skilling side, we joined the ITU's Skilling Initiative earlier this year to support that. But if I could, I'd just like to come back on some of the things that have been said, because I think it's a fascinating conversation. And I think that what we've got, very interestingly, here at the UN is something like— I mean, look, I don't want to be presumptuous, but something like a consensus that adoption of AI could be good. And I think that, you know, we should focus on that question of how do we diffuse and how do we adopt and how does it happen? Happen in a responsible way, and how does it happen in an inclusive way? And from a business perspective, of course, you want it to be inclusive because that increases the size of the market. You know, if 15% of the world is, you know, has a disability, you want to find a way to address that. So, we've got to look at this together. And, you know, let me talk really quickly about responsible AI and about skilling. There's clearly, as has been shown here in the conversations at the UN a desire across the world for a framework to ensure that the technology is responsible and is adopted in a responsible way. And that mirrors the approach we've taken in Microsoft and many companies, you know, for a decade or more. You might— my kind of strong advice, and this would be what you'd hear from any private sector, speaker, I think, would be that the conversations around the new AI bodies should have a multi-stakeholder kind of lens to them. So, things like, for example, the scientific panel, which is being designed to equip the UN to remain relevant in the conversation. If that panel does not include people with an understanding of what the private sector is doing at the frontier of the technology, that creates a risk for the UN that the conversation is less relevant. So, multistakeholderism, you know, it's not just a sort of nice add-on. I think it's really crucial in that conversation around responsible AI. In turn, we will benefit, the private sector, if something like a global interoperable approach is adopted that allows us to address the entire global market, right? So, this is potentially a win-win from the diffusion perspective. And then, on the skilling side of things, the people who will benefit from this technology are not necessarily those who have, you know, first mover advantage in inventing it or adopting it. It's those who find the most creative and productive way of using it. And that does come down to the question of capacity building and skilling. And this is really, really critical, not only in terms of education, which is very important, important. But all the way through, right up to, you know, the senior leaders of boards of companies, if people are not thinking now strategically about how to adopt AI, they risk being overtaken by those who are. And so, driving that capacity together through a multi-stakeholder approach, through these skilling programs, is going to be really important.

Thank you very much, Chris. And thank you for mentioning our AI Skills Coalition. And thank you for being a partner there, indeed. So, now is the time to turn to you. And to hear perspectives of member states and other stakeholders here on what you see as barriers to global AI adoption and also how to overcome them. We have quite a few people signed up. I'm told we'll limit the interventions by 2 minutes. It'll be 2-minute interventions. My job will be made very easy because your microphones will be cut off after 2 minutes. I don't need to do anything. So just please— and of course, if you don't need the whole 2 minutes, feel free not to use them, because I think other colleagues going after you will be very grateful to you for that. So we'll start with Zimbabwe, please.

Thank you for giving me the floor. Artificial intelligence, we've heard this afternoon, is advancing at a very extraordinary pace, transforming the way we live, the way we work, and solving—

the way we solve problems. It clearly holds great promise for accelerating progress on Sustainable Development Goals. So we welcome this exchange of ideas on how we can effectively leverage AI for inclusive development. But what is also clear is that we should tailor this technology to fulfill our needs. At the same time, we must confront the very real challenge of the global technology gap. Many developing countries, including mine, Zimbabwe, do not yet possess the full capacity to harness AI's transformative potential. This has been said by many in this session. Therefore, bridging this divide must be central to our global conversation on AI. As we go forward. To this end, Zimbabwe emphasizes the importance of capacity building, knowledge sharing, and equitable technology transfer. With strengthened capacity, it can also foster momentous change in critical areas such as healthcare, education, disaster preparedness, social protection, and climate action. It can help extend essential services to underserved and vulnerable populations, truly leaving no one behind. However, as we embrace the opportunities, we've also heard today that we must also remain vigilant to the risks. The misuse of AI through deepfakes, disinformation, and so forth can be quite harmful to our societies. This underscores the need for robust governance frameworks and effective human oversight grounded in ethical standards and shared accountability. With inclusive cooperation and responsive— responsible innovation, we can ensure that the benefits of AI are shared equitably and that the risks are mitigated. I thank you.

Thank you very much. Microphone to Costa Rica.

Muchas gracias, señor moderador.

Thank you very much, Mr. Moderator. Costa Rica wishes to highlight three points. First, we must spread what already works. Community-level AI already predicts floods, balances off-grid solar energy, and spots crop pests. These tools can reach every classroom, clinic, and farm by 2020. If we share data that is privacy protected, fund open-source models as digital public goods, and pool finance between the Global North and the Global South. Second, innovation must go hand in hand with governance. Generative AI magnifies bias, disinformation, and exclusion. Interoperable transparency standards, data stewardship, and accountability building on the Global Digital Compact UNESCO's ethical recommendation and the UN Principles on Information Integrity will give entrepreneurs certainty while protecting rights, especially where regulation is still lacking. Coordinated governance also unlocks cross-border capital flows and talent flows, which are vital for scale. Most importantly, international standards lay the groundwork for member states to hold advanced AI companies accountable, as stable economic growth removes the need for risky R&D and premature deployment. Third, we must improve things for the global majority. Nearly half of the world still lacks reliable connectivity, affordable computing abilities, or advanced digital skills. Targeted investment in resilient networks, regional research hubs, community data trusts, and inclusive training, especially for women and youth, will let local innovators address local problems problems and join global AI markets. Blended finance instruments can bring in private capital, while South-South knowledge exchange can accelerate adoption. Measuring impact through inclusive indicators— the speaker's microphone was cut off.

Thank you very much, but it's testing the system, so sorry that we're starting with you. So then next to Singapore.

Thank you.

Singapore recognizes AI's potential to address the challenges of our time and enhance our collective economic and social potential. We also acknowledge the risks that come with AI. As such, Singapore has sought to build a healthy AI ecosystem with enabling policies and frameworks that serve as systemic guides for safe and responsible AI development. As trust and safety underwrite confidence in Singapore's AI landscape, landscape, we have taken a multi-pronged approach to protect public interests and build consumer trust while ensuring space for governance-enabled innovation. For instance, we develop practical testing tools such as one called AI Verify that help organizations validate the performance of their AI systems against internationally recognized AI governance principles. We have also developed sector-specific tools such as in finance and healthcare that support efforts by our companies and researchers to develop new AI and build products and services safely and responsibly. To supplement these tools, we have implemented laws and regulations to address AI-related harms and risks, which are regularly reviewed to ensure that they remain appropriate and fit for purpose. Second, it is essential that we commit to enhancing digital inclusion so as to enable and support innovation. In this regard, Singapore has implemented a suite of initiatives to enable digital inclusion across the board. For instance, our Digital Skills for Life framework equips Singaporeans with everyday digital skills, including through utilising GenAI modules to help less digital citizens learn to use AI tools safely and responsibly. Third, we also seek to enhance international partnerships, to enhance interoperability, and to develop global standards and best practices for responsible AI. It is with this in mind that we created an AI playbook for small states. This playbook, developed in collaboration with partners from the Forum of Small States such as Rwanda, is an anthology of best practices addressing the unique challenges faced by small states, including constraints in resources, data, and talent. It also serves as a resource for small states as they seek to play an active role in advancing the—

Thank you very much, Singapore. Nepal.

Thank you. Thank you.

AI is transforming and reshaping our world. Its impact is profound and its potential unbound.

While AI's potential is immense, it also brings significant challenges, including bias, lack of transparency, data privacy risks, ethical complexities, job displacement, and impact on societal and cultural values. Many developing countries and LDCs face enormous challenges ranging from gaps in policy, skills, and knowledge to limitations in capacity, resources, and infrastructure. Additionally, issues of accessibility and affordability further compound their challenges because without accessibility, AI is also losing its market and enterprising and innovation. It is therefore crucial that these principles such as inclusivity, equity, fairness, and ethical integrity are at the core of AI implementation and governance so that AI benefits everyone. So in this context, first, we must close the digital divide divide and AI divide by helping create digital infrastructures and accelerate technology transfer, knowledge sharing, and capacity building through training, digital and AI literacy, and education, among others.

Second, we should promote entrepreneurship and prioritize investing in AI startups in developing countries. Third, adequate and predictable financing for AI readiness and building AI ecosystems must be prioritized.

Fourth, we must ensure multi-stakeholder collaboration across governments, international organizations, academia, and private sector to harness AI's full potential. To conclude, Nepal, which recently unveiled its National Artificial Intelligence Policy, is ready to foster a robust partnership with the international community and work together where AI serves the progress for all.

Thank you.

Thank you very much, Herminia.

Thank you, Thomas.

This special meeting is indeed timely and highly relevant. We, as we acknowledge the importance of harnessing AI for the implementation of SDGs, strategies. The PACT for the Future underscores the need to inclusively assess and address the potential impact, opportunities, and risks of AI systems on sustainable development, well-being, and rights of individuals. In this regard, there is a need to focus on how we can unlock the potential of AI to generate additional opportunities for sustainable development and assist the developing countries, in particular countries in special situations including landlocked developing countries on equitable access to the AI. We are in a view that the global action should result in making the development of AI predictable and comprehensible. It is also important to ensure that comparability of AI and the human rights, including economic and social rights. We should ensure that human decision-making remains central to the AI and put efforts to make AI accessible so that the cooperation between developing countries benefits from that. Empowerment of women and girls in the field of AI and their access to technologies and innovation should remain an important priority to enable their full, equal, and meaningful participation in economy. ARMY is committed to reduce by half the gender digital divide across generations by 2026 through accelerating meaningful access to digital technologies and universal digital literacy. To put the perspective of inclusive and accessible AI into action, in September 2023, AI was first introduced as a field of study in the Armenian public education system, starting with 23 high schools in 2025 and expanding to 40— 41 in 2026. The program ensures students in every region gain equal access to advanced AI coursework. We welcome the ongoing process to create independent international scientific panel on artificial intelligence and global digital— global dialogue on AI governments and ready to contribute to the global—

Thank you very much. Sorry for that. So, it will be UNESCO, followed by Algeria and the Save Our Environment 10-member group. UNESCO first.

Thank you, Thomas. At UNESCO, we view artificial intelligence as a transformative force that, if governed responsibly, can accelerate progress across the SDGs. UNESCO's mission is to ensure AI benefits humanity by addressing ethical, legal, cultural, and social impacts. In 2021, member states adopted the Recommendation on the Ethics of AI, the first global standards guiding responsible AI use and supporting our efforts towards education, reducing inequalities, and building strong institutions. That's why UNESCO chose the impact of AI on the media as this year's World Press Freedom Day theme. While AI brings powerful tools for journalism, it also raises urgent concerns: misinformation, transparency, and editorial independence. As AI's influence grows, so must our commitment to a free and pluralistic press. AI is transforming work, creating opportunities but requiring reskilling. UNESCO has helped over 10 countries develop AI competency frameworks and launch training in hundreds of institutions. Our Global Skills Academy makes free AI courses available to the 360 million out-of-schools and unemployed youth worldwide. UNESCO focuses on how AI can strengthen justice systems. We have trained more than 36,000 judicial actors in 160 countries on freedom of expression, access to information, and journalism, over 8,000 of whom have joined our program on AI and humanity— on human rights. To meet rising demand, UNESCO developed a tool, the Readiness Assessment Methodology, now used in over 70 countries to support AI development. UNESCO's AI work powers real solutions, from early disaster warning to preserving cultural heritage. Partnerships are essential. Through initiatives like the UNESCO Business Council, we're building strong collaboration. The upcoming Global Forum on the Ethics of AI in Bangkok, supported by ITU and UNDP, will be a key platform for advancing dialogue and implementing—

Sorry, UNESCO. So I think I stole a few seconds from you in the beginning, so apologies for that. So it'll be Algeria, followed by STI Forum member group and Türkiye. Algeria first.

Thank you for giving me the floor, and thank you for organizing this timely meeting and keeping AI as a priority for our Council. We also extend our gratitude to the speakers and panelists for their valuable remarks. Our discussion here is an essential contribution to steering AI in the right direction, one that leaves no one Regarding today's topics, I would like to express the following points. First, on the barriers faced by the developing countries, as covered by the panelists, they encompass technological, economic, social, and governance aspects, including lack of investments and expertise, limited infrastructure, lack of regulatory frameworks and strategies, data deficits, Second, international cooperation, partnership, and capacity building are essential to overcome these barriers and allow developing countries to be equal partners in global AI governance discussions and in harnessing its potential in advancing SDGs. The key priorities are establishing inclusive international cooperation mechanisms anchored in international law and the UN Charter, further, building human capacities and mobilizing financial and technical resources for infrastructure development. My third and final point, digital divide. Over half of Africa's population, 600 million people, still lack reliable access to electricity, which is a technology from over centuries ago. We don't want to see once again Africa and other regions prevented from enjoying technological benefits and opportunities, and AI's potential decisive impact on healthcare, education, and economic growth. Unless we confront this stark reality, artificial intelligence will only represent for humanity another face of gaps, divides, and inequalities. We must remain committed to ensure that AI—

Algeria. Now, STF Forum 10-member group, all the way, Turkey, Indonesia. STF Forum 10-member group.

Thank you.

Good afternoon, everyone.

I'm representing a 10-member group, a group of 10 experts with different fields, and we have consensus on that, that this topic is affecting all of our fields and across all SDGs, that there's not a single one that is not affected by AI currently. We are monitoring closely closely the development of the technology and the effects that it has, and this event is a very good step in that direction. Of course, science is very important in the whole thing, and we believe that there will be a few next important steps like fusion of AI and biotech, which will bring real revolution globally. Of course, There are positive and negative effects on SDGs. We heard this morning at the beginning of the event, in the introduction by Mr. Singh Gill, about the medicine breakthroughs. Also, agriculture has revolution due to AI. But also, we have negative examples. There is huge increase in energy demand.

Big companies are investing now in dedicated nuclear power plants, for the data centres that they are running. It goes with environmental impact, unavoidably.

And of course, divides that we are talking about, where we have a clear need for capacity building. The fact that we have 4.7 billion people on the planet with smartphones, 87% of all the phones, gives us possibility to bridge this and, for example, to give education opportunities through AI recommendation systems globally. Jeffrey Hinton—

Sorry, sorry, I think the time is up now. So now it will be Türkiye, followed by Indonesia and Uruguay. Türkiye, the floor is yours.

Thank you, Mr. Moderator. First of all, I would like to thank the President of His Excellency Bob Rae, for the organization of this special meeting at a timely moment, as we are at the same time discussing the future modalities of AI governance within the United Nations. I would also like to thank distinguished panelists and speakers for their remarks. In today's globalized world, AI is a strategic enabler for achieving the 2030 Agenda and its SDGs. However, irresponsible development and use of artificial intelligence may lead to irreversible harm to our children, our societies, and our economies. Therefore, the protection of ethical principles, human rights, and the rule of law is of utmost importance in the development of AI-based technologies. Distinguished participants, Turkey adopted a national artificial intelligence strategy and is applying AI in key public service domains. We promote broad digital inclusion through national AI literacy and digital skills training programs in primary education, vocational schools, and the public sector. Dear colleagues, capacity development gaps, especially in the field of AI, pose many challenges for developing countries, and in particular LDCs. The UN's Technology Bank for LDCs, which has been proudly hosted by Türkiye, has been playing an instrumental role in supporting the technological and innovative capacities of the LDCs. The bank has a critical role to play in bridging the digital gap between the LDCs and the rest of the world. We cannot allow the world to live in 2 different centuries. We would like to underline the importance of utilization of the UN technology.

It will be Indonesia, followed by Uruguay and Brazil. Indonesia.

Thank you. Indonesia shares the view that AI can serve as a powerful enabler of sustainable development. However, as many have said, developing countries—

the promise—

for many developing countries, the promise of AI remains constrained by structural and systemic barriers. First, on the digital and data divide rooted in limited digital infrastructure, insufficient data governance frameworks, and a lack of AI-ready human capital. We believe that the well-targeted investment in AI literacy and inclusive digital ecosystem, particularly for developing countries, are crucial. Second, the absence of harmonized global architecture structure that limits the capacity of the Global South to meaningfully participate in the benefit and benefit from the AI, and that exacerbates inequality in access. To address this, Indonesia looks forward to the global— to the establishment of AI scientific panel and global dialogue on AI governance. On our part, as we acknowledge the fast-growing of AI roles in our socioeconomic fabrics, Indonesia Indonesia itself has taken concrete steps to anticipate the dynamics. Over the past years, we have launched our national strategy on AI, established Indonesia AI Innovation Research Centers, and initiated AI skills training for youth. We also believe that strengthened international cooperation is necessary to ensure a development-oriented digital transformation. And to conclude, we share the importance of cross-regional to strengthen policy coherence and emphasize the urgent needs for robust and inclusive AI governance. Thank you.

Thank you very much. It will be Uruguay, followed by Brazil and Venezuela. Uruguay, please.

Muchas gracias.

Thank you very much. Limited access to quality data, dependence on external technology, specialized talent, and computing capacity are some of the structural barriers that developing countries face in order to really harness the potential of artificial intelligence to achieve the SDGs. In order to overcome these asymmetries, developing countries need to build inclusive AI ecosystems that are tailored to the local context and that require collaboration between governments, multilateral organizations, the private sector, and civil society in order to provide digital skills, to facilitate technology transfer, and to promote international cooperation. The National AI Strategy of Uruguay is in line with this approach, promoting alliances and applied research, technical and ethical training, and regional integration as driving forces for our own solutions. We're encouraging solutions such as open laboratories that enable us to co-create technology with social participation, with ethical guidance, in order to deliver access to artificial intelligence and to ensure participation in the digital economy. However, regulatory fragmentation is a concrete stumbling block, which means that we need to adopt flexible and robust regulatory frameworks that enable interoperability, open data, and shared digital spaces. And that's why we support the UN in building artificial intelligence global governance through the panel of independent experts mentioned in the Global Path for the Future. And to conclude, a question in English.

You could mention effective experiences for strengthening human and technical capacities in artificial intelligence in developing countries, and how can they be scaled up in the regional level to advance the SDGs? Something was mentioned, I think, from Microsoft regarding—

Thank you very much, Uruguay. We got the question, I think. So then will be Brazil, followed by Venezuela and Switzerland.

Thank you.

Thank you.

Excellencies, colleagues, we appreciate this timely meeting and the insightful comments from the panels.

Mindful of time, I will only highlight a few guiding principles essential for Brazil when discussing AI in this venue. The United Nations must remain the central forum for AI governance to ensure the participation of perspectives from multiple stakeholders within their respective roles and responsibilities, and in particular stakeholders from the Global South, and to ensure that their perspectives shape policies that affect us all. This is essential for governance legitimacy and efficacy.

While embracing global coordination, we must affirm each, each nation's digital sovereignty. At the same time, infrastructure investments are fundamental for equitable AI access.

We must bridge the digital divides while creating conditions for technology transfer that builds local capacity, not dependence.

Open ecosystems should be prioritized through open-source development and knowledge sharing to minimize reliance on proprietary systems that concentrate power. Barriers to participation contradict our commitment to the SDGs.

Collaborative research must empower local technological capacities by recognizing diverse expertise and creating space for co-creation. AI's impact will depend on our commitment to equity, sovereignty, and meaningful collaboration.

Thank you.

Venezuela, followed by Switzerland and Sweden.

Venezuela.

A very good afternoon. Thank you very much. Thank you very much. Venezuela is grateful for this meeting where we're talking about something that defines our presence and that will change our future. Artificial intelligence is not only a disruptive technology, it is a catalyzer for global progress. Key sectors like health, education, agriculture, and justice— in those sectors, artificial intelligence is making a significant difference. However, we cannot ignore a critical challenge, which is the technology gap. While developed countries are making strides forward with artificial intelligence, a lot of nations in the Global South face barriers such as a lack of infrastructure, lack of access to quality internet, and skills for specialized workers. This disparity does not only limit these countries' growth but also entrenches global inequality. Venezuela calls for increasing international cooperation. At the same time, we call for more inclusive policies, and a concrete example of this is scientific development licenses. Cutting-edge ones must be accessible, and for them to be accessible, AI needs to adapt to local contexts respecting cultural diversity and linguistic diversity. We must also increase investment in the Global South. These countries in the Global South have key resources— energy, basic resources, raw materials— and a young population that, combined with artificial intelligence, can transform their economies. It's important to underscore that artificial intelligence must be used for peaceful purposes and in line with the spirit and principles of the UN Charter. The speaker's microphone was cut Thank you very much.

So now it will be Switzerland followed by Sweden and Nicaragua. Switzerland first.

Thank you very much, sir. First of all, I would like to thank the speakers for their stimulating statements. What I take away from this discussion is a key word, partnership. Partnerships which are strong and inclusive to help us implement the potential of artificial intelligence. We support such cooperation very strongly. And let me give you 2 specific examples from our Swiss experience. First, the partnership between the Federal Polytechnic School in Zurich and the United Nations, which aims at strengthening links between science and public policy, including in the area of artificial intelligence. This cooperation model will be illustrated during a side event on the of May on the margin of SDI Forum will be co-chaired by Switzerland, Finland, Singapore, and the Polytechnic School. Secondly, the ICANN initiative is an example which provides the resources of computing, data, and competencies in AI to accelerate those projects which are related to Sustainable Development Goals. These platforms show how responsible innovation can be catalyzed by multi-actor alliances that that you've just mentioned. We promise that we'll have a balanced approach to AI governance, which will link encouragement, innovation, and a solid ethical framework. We believe in a multi-party process— processes and the central place of Geneva with its international organizations and experts to fashion an inclusive digital future for us all. I thank you.

Thank you very much. Sweden, followed by Nicaragua and Saudi Arabia. Adrian, first.

Thank you very much.

The conversation today we've had has deepened our understanding of the key barriers that prevent developing countries from leveraging AI for economic transformation. And in our conclusion, the first one is that a significant gap in digital infrastructure, such as reliable internet access and advanced computing resources, fundamental for AI deployment. A second observation is that the regulatory and policy challenges, such as the fragmented governance and lack of standardized AI policies. It hinders interoperability and seamless integration of AI technologies. To overcome these barriers, governments, international organizations, academia, and the private sector must work together to create the robust AI ecosystems that we all want. It could be joint research initiatives, capacity-building programs, sharing best practices and resources. Fostering ecosystems for innovation and entrepreneurship are key to accelerate the progress towards the Sustainable Development Goals. Startups and innovative ventures drive economic growth and the global solutions, and my country is always ready to help lead the promotion and deployment of such ecosystems. We were therefore very happy to co-facilitate the Global Digital Compact last year, and now it's It's all about implementation of it.

So by addressing infrastructure gap and enhancing collaboration, we can create an inclusive digital future for all.

Thank you very much.

Thank you very much. It will be Nicaragua, followed by Saudi Arabia and Russian Federation. Nicaragua first.

Muchas gracias.

Thank you very much. I think if we want to lay solid foundations for AI in the Global South, first of all, we must recognize the real barriers that we face. Not only the lack of infrastructure and capacities, but also restrictions such as unilateral coercive measures that include exclude us from fairly accessing the digital economy. AI must be a public good that overcomes inequalities, and that's why we must work together to establish ecosystems that really work to work in our context, that are sustainable and that are. Are designed to meet our own needs. From our experience, South-South cooperation and regional alliances can make a real difference, especially when it comes to sharing knowledge, strengthening systems, and building public digital solutions. It's also key to have clear rules that avoid fragmentation and that ensure that we all have a voice on how these technologies are regulated. Thank you very much.

Thank you very much. Saudi Arabia, followed by Russian Federation and China. I will have to regretfully close after China. So Saudi Arabia.

In the name of God, the merciful, the compassionate, I'm going to briefly mention some of the most recent developments we've had in the Kingdom of Saudi Arabia to drive forward innovation. My country hosted the— a conference in February which was the most important tech event bringing together stakeholders in the tech sector, confirming our interest in driving forward digital investments. At this conference in February, a lot of investment was announced. We've announced millions of dollars to make the Kingdom a driving force in building our digital future. We've used these investments to enable us to boost the digital capacities of our citizens, our institutions, and of our region. We have also encouraged startups— Can you repeat the question? We are the biggest digital economy in the Middle East, and we have also declared investments of a billion dollars to build a technology regime based on artificial intelligence and robotics. We've also established a regional center by Lenovo in Riyadh, and Google has also announced new business in the kingdom to help meet regional and global. Needs. The Kingdom still aspires to— always aspires to share its experience with other countries and stakeholders in this area to ensure that we leave nobody, no individual or country behind. Thank you very much.

Finished the interpretation in my earphone, so thank you very much. So now will be Russian Federation.

Distinguished colleagues, Madam Thank you, Mr. Moderator. The meeting today has shown that in order to overcome the digital divide is one of the key tasks the UN is facing. Foundation for such work for us are the decisions that came out of the World Summit of Information Society. They remain not fully implemented, and we advocate continuing the WISIS program for another 10 years. UN has to play a central role in regulating AI. Attempts to discuss this issue outside of the UN in closed private formats are counterproductive. We note in this regard a very important role played here by the ITU, the role they play in developing fair standards and requirements on ICT. We are convinced that developers, and it was already mentioned, has to strictly abide by the UN Charter, the legislation of the country where the ICT is being used, and also respect the sovereignty of states. AI has to help with progress in attaining SDGs, and we're proposing the following steps for that. First, supporting the systematic development of initiatives on using AI at the local level, including computer training organized with the participation of government, business, and civil society. Second, help create international consortium for development purposes with the participation of developing countries. And third, help develop the human potential of developing states. 2025, 2030, institutions of higher learning in Russia will introduce their own educational programs for developers, analytics, and researchers focusing on artificial intelligence and the private businesses. Private sector will provide at the very least 30% of the funding.

Thank you very much. Now will be China. Now, the good news, I've been told that we will have— we'll be able to have 2 more interventions after China. So, we'll have China followed by Sovereign Order of Malta, International Association of University Professors and Lecturers. So, China first.

Thank you. As a strategic technology driving a new round of scientific and industrial revolution, AI is transforming the way we live and work. In 2023, Chinese President Xi Jinping put forward global AI governance initiative as China's proposal on the topic of global AI governance. International cooperation on capacity building is a key component of global AI governance. Last year, UN General Assembly adopted by consensus a resolution proposed by China enhancing international cooperation on AI capacity building. China launched the AI Capacity Building Action Plan for Good and for All and proposed the establishment of a group of friends with the purpose of implementing the UN resolution and to follow up on the Global Digital Compact. And China is a proponent and a practitioner of international cooperation since these establishment of the group. China has organized 2 seminars and has achieved great results. Last month, China and Zambia's permanent missions distributed a questionnaire on the Group of Friends, and we have received a lot of feedback. And China, based on the feedback, will organize practical cooperation. And this morning, China and Zambia's permanent missions, together with China Association for Science and Technology, co-hosted a meeting of the Group of Friends with positive response. And China will continue to uphold the principle of promoting AI for good and advancing the development of AI to enhance the common well-being of humanity. We will remain committed to fairness and inclusiveness, respecting the equal rights of developing countries to develop and benefit from AI. China will support coordinated global governance and the leading role of the UN in advancing the international governance of AI. Thank you.

Thank you very much. Sovereign Order of Malta. Thank you, Moderator. As the international community works to build inclusive AI ecosystems, we must remember that the questions we face are not only technical but moral. The Sovereign Order of Malta remains steadfast in its advocacy concerning the profound and far-reaching implications of AI. In this spirit, our mission hosted a side event this past February to illuminate the ethical and moral challenges that accompany the development of AI, and we take this opportunity to recall the reflections of Father Paolo Benanti, advisor to the late Pope Francis and one of our speakers, who reminded us of a simple yet profound truth: technology is neutral. He recalled how parkway bridges in mid-century New York were built too low for buses to pass, a subtle act of design that effectively screened out low-income riders. The point was clear: infrastructure encodes power, and today it's not bridges but algorithms that determine who's included, who's visible, and who's left behind. This raises a moral question for all of us: who builds these systems? Whose intentions do they reflect? And who is accountable when they fail? The Order of Malta approaches these questions through the lens of humanitarian practice. Although we may not build AI systems directly, we strongly advocate for those who do to ensure that the voices, needs, and dignity of the vulnerable are included in their design. When training datasets lack representation, algorithmic tools may exclude and/or mischaracterize the very communities they are meant to serve, reinforcing rather than reducing existing inequalities. AI-driven decision-making in healthcare, for instance, can overlook rural populations with insufficient health data, exacerbating disparities in diagnosis and treatment. There is also, however, real promise in the Global South and beyond. AI can help bridge longstanding development gaps if designed and deployed with equity in mind. We believe that technology should be a bridge, not a barrier, not a Let us ensure AI is guided by solidarity, not supremacy.

I thank you.

Association of University Professors and Lecturers.

L'Association Internationale des Professeurs et Maîtres de Conférences des Universités se félicite.

We welcome the progress accomplished by UNESCO in the area of artificial intelligence. The implementation of AI recommendations is an example of success. This has revolutionized several sectors. We are thinking that an additional document to the Pact for the Future is absolutely necessary to cover ICT and AI. We're not trying to multiply additional instruments, but to create an original document so as to have a convergence between technology and society where everyone will find their place for the purposes of sustainable development for humankind. I thank you.

Thank you very much. So I really appreciate everyone's interest and really high activity. So really apologies to those who are not able to make it due to this interest and very limited time that we have. So however, you're invited to send your written statements to e-statements@ e-statements@un.org, e-statements, one word, @un.org, and will be published. Now, just coming back to our panelists, and literally we have very limited time, so I will ask you to probably say one sentence of your key takeaway, and that will include answers to the Uruguay question as well. Just to mix it up a little bit, Chris, I'll start with you. I think you're quick on your feet, and so Chris, and then we'll go to Omayyama And Gerry, so please, Chris.

Well, look, thanks for including me. Thanks to Ambassador Ray for his leadership on all of this work in the committee. To Yorgos, really quickly, we think about the cloud as being in the sky. Like, it's not. It's physical infrastructure. And so, considering questions of infrastructure in regional— with a regional perspective makes sense because, you know, you need that contiguity and you need the connectivity through the region. So, I do think infrastructure could be good thing to think about. But I guess my main point would be, I hope that we demonstrate in these conversations the value of multistakeholderism. We stand ready to continue to contribute to those.

Thank you. Thank you very much, Chris. Omayyad?

Sure. I think I'll add on to that. And in one sentence, I would say that the point is not just to build AI. It's about building the right systems together with the right people, solving the right problems.

Thank you. Thank you very much. Jerry?

If there is one thing that we learned from the COVID pandemic, it is the fact that no one is an island unto himself. SDG 17 talks about constructive partnerships. It is important that we work closely together, South-South cooperation, growing AI together, and making sure that we are responsible in its use and we are leaving no one behind.

Thank you.

Thank you very much. Indeed, for me, just to close, I think the 3 words or the 3 concepts I carried here is, First is importance of local ingenuity in action. So, I think we talked a lot about how the AI needs to be locally developed, and that's really crucial. We talked a lot about partnerships, indeed, between private and public sectors, academia. We also— civil society that we heard today about, you know, between different regions as well, and also about global solidarity to make sure that no one is excluded and we don't create a new, and this time, AI divide. So, indeed, these are the for me, important takeaways. I would really be a bad ITU representative if I wouldn't be inviting you now to see you at the AI for Good Global Summit from July 8th to 11th in Geneva. We'll be really continuing these conversations with a great multistakeholder group there. Around 10,000 people are expected there, so I hope to see you all as well. Again, big thanks to Ambassador Bob Rae for having this meeting, for inviting me here as well to facilitate these conversations. Thank you very much again, and join me Thank you very much.

Thank you. Thank you so much. Thank you. Thank you. It's my task now to simply wrap up. I'm not going to give a speech. I know this is a wonderful opportunity to sort of walk out and either sing or talk, but I will just talk. Just to say to all of you, thank you so much, and thank you for staying. I think we've really tried to have a stimulating overview of what are some critical questions around the governance of AI, the uses of AI, the things we need to do to make AI much more accessible to everyone. And I think we've had a good review of all this. I want to thank the panelists very much. I want to thank them for coming. Many have had to travel a distance to get here. And I really want to thank all of you for doing this. I want to thank the moderators for keeping us on on an even keel and for allowing for as many speakers to speak as possible. The last thought I want to leave with you is this: we all know that the UN is going through a rough time, and we know as well that there are lots of discussions about what the future holds for the United Nations and also what What are the challenges that we face? And there are many detractors of the organization out there, only too eager to speak negatively of what the UN is doing and has to do. I just want to say to all of you, it is not possible for the UN to fail for the simple reason that there's too much for all of us to do together, and the reality is we can't do it unless we do it together. And that is the thrust of everything that's been talked about today. As has been said, partnerships, regional cooperation, building infrastructure in ways that will reach billions of people, developing rural developing ways of governing AI in such a way that it does not become a destructive force but remains something which allows people to learn more and have access to a better life. None of these things can be done without the networks that the United Nations has created since 1945. I think we need to understand that, and that just as we are going through a process of reconsidering how we can structure our work better, let us be— there be no doubt whatsoever that there is work to be done. There is work to be done. And this is something which I think has to animate our sense of purpose as we look at budgets, as we look at ways forward, as we try to to winnow out some of the political arguments and come back to the foundation that the UN was created, yes, to reduce conflict, but it was also created to allow people to develop in freedom and to develop in ways that will allow, as we say now, no one to be left behind. But the mission of solidarity is still such vital mission of the United Nations and all of the agencies and all of the people that are working. Some people see only duplication. I have to say I disagree. The fact that we have many centres of excellence within the UN and around the world that are dealing with AI is not a problem. The need for us is to bring it together and to weave it together, not to create uniformity, not to have one centre central driving system, because that's impossible. That's not going to happen. But to have a way of creating the means of cooperation and the means of consolidating what we know and making sure that we're taking the best advantage of what we have. And I'm delighted that the UN family has been able to participate so fully in the discussion, the agencies that have been involved and will continue to be involved. And yes, Thomas, we will be coming to the— I will be coming to the ITU Summit, and I hope there will be many others who will be coming. It's critically important that we keep this conversation alive, and I'm delighted, really delighted, that ECOSOC has been able to play a constructive role in this conversation. So thank you all very much. Muchas gracias. Merci beaucoup. Thank you. Take care.