 Welcome to the AI for Good Global Summit. Our next guest is Thomas Lamanouskas, Deputy Secretary General of the ITU. Thomas, thank you so much for joining us. Thank you very much, Jillian. It's really a pleasure to be here. Thank you. Let me start by asking you, what would you like to be the outcome of this global summit? No, so thank you very much. I think some of the outcomes we're already achieving is really bringing global community from different stakeholders together here to look into the AI as a concept, as a solution, as a phenomenon, but also as a very practical application, what it is in the future, with this conceptually, and what it is already today. So I think, because that awareness building and inclusive dialogue is already very important by itself, and I think we're really having here with a lot of people. But then, of course, the second piece is I think also to start thinking about these issues that people are talking about, so regulation and governance aspects of the AI to make sure that the AI is not just that's really for good and that kind of things that the AI can deliver to everyone is also not overshadowed by the things that the AI can threaten us in different ways, shapes and form. But still, and then maybe the third piece of that is also, like again, reaffirming and reasserting that piece of for good. And understanding, sometimes in this dialogue of today, we're kind of losing the fact that the AI is in the end of the day as every technology can serve good but can have risks. But sometimes now the dialogue with the risks overshadows that dialogue of the good solutions. And I think with being here, being immersed here, seeing the real solutions, seeing people talking about real solutions, seeing them on the ground with a lot of robots, I think another piece is, I think allows to kind of reconnect to the fact that in the end of the day, this technology has enormous potential for sustainable development agenda and for our everyday lives. And I think that's what for me, if we achieve that, that both governance dialogue, reaffirming and reasserting and really seeing in the right ways how it can be used for good, but then doing that in this very cohesive, joined up way, I think that is a great outcome. And I think we have ingredients here for that. We have UN agencies through different formats. And after this interview, I'll be running to the UN round table. And yesterday, we had a preparatory meeting of the interagency working group on AI of United Nations. It's our technical term with the UN family as experts are coming together to discuss how jointly to deal with both good and risky sides of the AI. We have different stakeholders. But also what's important, we have this visual representation. If you look at the exhibition floor, look at the things, you'll see those robots squaring around you. You have different solutions there. You have artists in our stages kind of really immersing you in what we now call here artistic intelligence. And I think that also gives you the feel that that is real and that touch to what is real today. And I hope that also, and it's also a big output of the summit like that rather than just a discussion, the first round table is that you actually connect those global general discussions with the reality with actually happening on the ground. And I think that is super important for us. Okay, so you talk about it being real. How can the AI further take on these SDGs? How can they advance them in reality? No, so I think first of all, there's a lot of very practical solutions here. Like both here and in our other work. As you released in earlier this year, there are the new addition of what we call Compendium of AI Solutions for UN. So actually we did that together with the UN family, other 40 agencies we collected 280 plus examples of AI use in the UN system. And 85% of those examples are directly related to sustainable development goals. They are like directly helping somehow advance sustainable development goals. For example, UNEP, United Nations Environment Program, working with the United Nations Office for Coordination of Commentary Assistance, on really looking at the weather patterns and migration patterns and predicting how the climate change and specific climate patterns change might impact future migration patterns and how that could affect humanity. Now, for example, World Food Program, working with different partners, looking into disaster management. For example, assessing from satellite imagery how the buildings are damaged after the disaster and whether the help is needed. We as ITU also have a state of initiatives. For example, we're launching actually today our global initiative on AI together with World Health Organization, WHO, to actually gain advance to different areas, aspects how we can use better AI in the health system to predicting the, both predicting diseases and helping doctors and so on and so forth. We're looking in AI, for example, in agriculture with Food and Agriculture Organization has gained through the ITU frameworks and so we call it focus group on that. We're looking in AI for disaster management and so another focus group on that. So there's a lot of like specific and real life examples there. When you look, you know, put your imagination and listen here, you see how many more solutions are there. For example, today was a presentation where the people were talking about using AI for forest fire prediction and flows and how to manage but also how to predict where the smoke will go. And this is now, regrettably, some of those things becoming everyday occurrence for a lot of people in the places where we didn't expect that to happen. And again, every tool, and AI is a powerful tool kit to help us at least to manage that somehow. So that is the thing for me, the specific examples show that we really need not, you know, to think beyond like overarching concept of AI into like specific solutions where AI, machine learning data can be really useful. Thank you. It's really good to hear some real examples. There is the risk, of course, that the technology isn't developed in an ethical way. How do we ensure that it is? I think this is a complicated question. And it's also still a very evolving question because I think we're still grappling with that. We have a number of legislative initiatives or policy initiatives around the world. Actually, we counted recently that around 96 countries, at least in our account, have some sort of initiative policy or legislative to kind of deal with AI. So there's a lot of activity there but I think there's a lot of still questions that need to be answered. There's also an UN system. We have our friends in UNESCO that adopted the ethical AI guidelines we have as a UN system. We agreed on ethical guidelines and principles of ethical use of AI within the UN system. So we're looking into that how to do it. Now, let's say some of the principles are emerging. Some of those principles include, for example, transparency and accountability. So that making sure that we kind of know what's happening at least to some extent possible in those. And there is some human control of the models. So then that includes avoiding of bias. The principles includes really ensuring fairness of the system. Includes security aspects of that. But also I think it's very important inclusiveness. And inclusiveness is along different aspects of that from inclusive data sets. And I'm again hearing complaints that, for example, data sets sometimes very biased in the terms of what they're used to. So depending where you live, those data sets and decisions can better or worse actually be helpful for you. But then also inclusiveness in terms of ability to use AI. And that's kind of where it comes even the basic thing which it uses. Look, remember, 2.7 billion is still not even online. And also let alone using AI. So those benefits are not there for them. And then, of course, capacity to use AI and capacity to protect themselves from risks. And also understanding AI models and understanding how AI works and helping it. That's also, it's an individual level but it's on a country level. So we still have a huge divide. And especially from that angle, from that divide angle, that I think that's the first very natural entry for the ATU because for 158 years we've been working to overcome those divides in the use of technology. So I think therefore today we still stand here and we are very proud to convene this event as a way to really bring the benefits of AI to everyone in an inclusive fashion and enable everyone to kind of manage those risks as well. Thank you so much for your insight. It's been fascinating. Thomas Lamanoskas, thank you. Thank you very much. More to come on the AI for Good Global Summit here in Geneva.