Media stakeout - Global Dialogue on AI Governance

Good morning. Welcome to this press stakeout on the Global Dialogue on AI Governance. I am going to give you just a few information about the rundown. We will hear from the President of the General Assembly, Annalena Baerbock, first, and then from the two co-chairs of the Global Dialogue, Griselda López and Reine Tamsar. They will give you a few introductory remarks, and then we will take a few questions on the dialogue. Thank you. So, I'll start by giving the floor to the President of the General Assembly, Mrs. Baerbock. Please.

Good morning, good afternoon from my side. To those of you who haven't been in the opening tomorrow, first of all, thank you so much to the co-chairs of this global dialogue. This is the work for almost a year. And behind the scenes, as always, more important than what you see afterwards on stage. But you two make it possible. Thank you so much. As we underlined all together this morning, artificial intelligence is reshaping how we live, how we work, and how we engage as humans with one another. And as the co-chair of the scientific panel underlined, intelligence always comes with power. Therefore, something with such power as artificial intelligence, with such profound implications for our economies, our social systems, our defense, and therefore our peace and security, but especially in our homes, even in our children's bedroom, can only be meaningful and safely if managed together. And the good thing: the UN, the United Nations, was built for moments like this. Even if the founders of the United Nations knew nothing about AI— they didn't even have a computer back then, 80 years ago— they recognized when drafting the UN Charter in San Francisco that global challenges with economic, social, and humanitarian character can only be solved together. That something with such power as AI needs global platforms in order to prevent that our social media platforms of our children are turned into battlefields. An inclusive AI governance is overdue. And as the Independent Scientific Panel on AI concluded in its final report, which we are discussing also here at the Global Dialogue, concrete next steps to close this gap exist, but each requires sustained investments in member state capacity to shape, evaluate, and deploy AI. And given the complexity of this technological revolution and its rapid pace, No country, no government, no company can nor should do that alone. We need governments, we need technology leaders, we need academia, we need civil society and international organizations working together to shape our digital future together. Therefore, this Global Dialogue is about giving us that same San Francisco moment, as the co-chair said this morning, of bringing everyone together to address a shared global challenge in the interest of all, leaving no one behind. I thank you.

Thank you very much, President of the General Assembly, Her Excellency Annalena Baerbock. I give now the floor to Her Excellency Griselda López from El Salvador, co-chair of the Global Dialogue.

Thank you. Good afternoon and thank you, thank you so much for being here. Madam President, thank you so much for your support. You appointed us, you gave us a task, a mandate that we have to fulfill that the members of the General Assembly decided in not only establishing the Global Dialogue, but also the panel as we saw, and we're very happy to be here to see these two important milestones be a reality. So thank you so much. And well, for El Salvador, as I mentioned at the beginning of the program, it has been truly an honor to serve as co-chair together with Estonia. It has been quite a journey. It's not easy to organize something for the first time, but it has been something at the same time really fascinating to be able to engage with a variety of stakeholders, as we saw here today, and also with many member states that were eager to finally have a say in artificial intelligence governance. And this is precisely what we saw today, that an inclusive and universal platform as the United Nations can make a difference, and it matters. It matters, I will say, more than ever with that San Francisco spirit that you say. But it's not only member states. Just before we left, we were listening also to a variety of stakeholders that were also wanting to take the floor, and we are definitely so much encouraged by their enthusiasm and their support. And that combination of member states, the scientific panel, private companies, academia, private sector, and so on— this combination, it's really important. We already have the scientific evidence, as we saw today from the scientific panel, and also it's really encouraging that this is just the first Global Dialogue, and we will see in New York the second iteration of this Global Dialogue. So rest assured that the work will continue, and I really hope, and I'm pretty sure, that it will be with the same universal approach and inclusive spirit as we have worked so far. Thank you once again, and continue supporting the Global Dialogue and continue supporting the UN in AI governance. Thank you very much.

Thank you very much, Mrs. López. Now I give the floor to His Excellency Reintam Saar from Estonia, the other co-chair of the dialogue.

Thank you. Thank you very much. Thank you for being with us here. Annalena Baerbock, many, many thanks for all the trust that you've put on us with my dear co-chair Ambassador López. It is a difficult moment for me because today it's my daughter who is turning 7 years old and she asked me, "Papa, why you're not here with me?" And I said, "Sorry, I have something important." So if this thing would not be important for me, I would not be here. Maybe this is evidence that we worked hard for 248 days to make this dialogue happen. So let me briefly say a few points. The first is on inclusivity and multi-stakeholder participation, because I think that this is the essence of this dialogue. So more than 4,000 participants, more than 1,500 written submissions, 24 side events, many, many stakeholders from around the world. So 1,800 entities. So this is what we call inclusive, inclusive dialogue. And we really— for us, it was so important from very beginning because we convene this dialogue, this universal platform, with convening power of the United Nations. So that's the difference of this dialogue to all the other dialogues that we have.. We want this to be inclusive so that every perspective, every voice is captured here. And of course, we'll have to write a summary about this report later on. So we really appreciate all participants coming here. And I met one of them, he said, "I just arrived from New Zealand." I said, "Wow." We thought that, you know, coming from New York is difficult, but from New Zealand, okay. So every region, every level of development, is represented here, and maybe it would be even better when the next dialogue will take place in New York in 2027, because we are setting the scene, so to say, for the next discussions. We can't do everything in 2 days because this was prescribed to us, yeah, but we really tried hard so that this dialogue would be inclusive. Secondly, maybe on risks versus opportunities, because this has been a huge issue for us from the very beginning when we started to meet with stakeholders, member states, and so on and so forth. So I mean, the risks are very real. And we heard also from representatives of a scientific panel that it's a dangerous business, yeah. But at the same time, I come from the country that has built a digital society, experimented, and try to extract maximum from this. And we are doing the same when it comes to AI. So I think that the best way to manage the risks is to do it, but also not to stop the innovation. We have to do it both at the same time somehow. And our kids, by the way, are doing— my older daughter, she's 14— she's 17. And she's sitting with her agent and doing her homework and everything. I'm a little bit frightened, but I see that she really likes her because this helps her to manage her time. So it's already there. We cannot stop this. We can manage this. And this is what we are trying to do here. And my final point is on human rights because this is a foundation for all of it. In our consultations around the world, I think that There was a very strong message that AI will only deliver on its promise if it respects international law and human rights. And this is not a brake for innovation. It is what makes people trust the technology. People need to know that AI serves them and not the other way around. So, and it respects their dignity, privacy, and freedoms. I think this is really important, and this was the message that we heard also from there.

Thank you. Thank you very much, Ambassador Tamsah. Let us now open the floor to questions, a few questions. Maybe I'll start with the host country's news agency, ATS. Laurent, can you please introduce yourself but also say to whom you're asking your question?

Sure, sure.

Laurent Siro, Swiss News Agency. First, happy birthday to your daughter. And then the question to President Baerbock, if I may. So in his speech this morning, the SG said that he was about to share his recommendation for a global fund on AI with the GA. So when should we expect— which horizon should we expect that fund to be operational? Maybe by the next dialogue in 2027? And what would be the amount targeted and the scope of of the Fund. Thank you.

Thank you very much for the question, and as the Secretary-General underlined, this is not the moment to watch, this is the moment to act. And as I said, the United Nations was built for cross-regional, transnational challenges, and this is one of the biggest challenges we probably have faced in the last decades. Knowing that we can only govern this together. And we have faced many other challenges. We'd always needed the cooperation between the private sector and governments, but in this dimension, this has been unseen. And therefore, yes, we need different parts. We need a governance structure, especially charting with our children. So he was also mentioning the regulations with with regard to protection of children. This is one part for the United Nations actions. And the other part is ensuring that we are not opening a new digital divide. So technology transfer, not only between countries, but also with regard to enabling the United Nations itself to use this technology, we are having debates to govern it around. So the issue about a fund for technology and also the use of it is something we have discussed within UNAID reform as well. And within the whole UNAID reform process, until the end of this session, which is beginning of September, we will take the next UNAID reform steps. The Fund is part of this bigger process, but obviously the Fund won't be set up until September, so it moves in the 81st session of the United Nations. So I hand over to my predecessor. The Secretary-General will hand over, underlining again that we are all a team at the United Nations. So, yes, there will be the next Global Dialogue in almost a year's time, and with this— these months to the upcoming next Global Dialogue, also the fund will be set up. But as we are starting it at the moment, I cannot speak about numbers, but it's crystal clear, —also another field—that in a world with billions and trillions, money is hardly the challenge. The main challenge is that it's spent in the benefit of all and enabling to bridge the divide and not deepen the digital divide we are facing currently. Thank you very much.

On this side, CCTV.

Thank you, Alexander. And I'm a correspondent of the China Media Group CCTV, and thank you for taking my question. My question is, why did the UN choose to convene the Global Dialogue on the AI Governance at this moment? And do you think the AI development of the AI is outpacing the global governance or the rulemaking or something like this? Thank you. Who would like to take this? For the President or the co-chair?

The co-chairs will take that out. Thank you. Xie xie. Thank you. Well, first, the Secretary-General also mentioned that in the PGA in their statements earlier this morning. This is not happening out of the blue. It's actually the response of years of deliberation of member states that began precisely with the Global Digital Compact, that AI was set precisely as one area that the UN should attention to it. And then last year, and this is full, full recognition to Spain and Costa Rica, they were the co-chairs for the resolution that established both the scientific panel and this global dialogue. And as mandated in that occasion, we're happy— we're already celebrating the first iteration of of this global dialogue. Here, back to back, as you already know, with different activities organized by ITU, like AI for Good and also the WISEs. And next year in New York, we're gonna have the second dialogue. What is going to happen in between, that depends on the next two co-chairs, that they're going to continue steering the conversation between this first dialogue and the second. What we hope, and my co-chair was mentioning, is to set really strong foundations on the conversation on AI within the UN. Co-chair, would you like to take the other part of—

Yeah, no, thank you. No, thank you. Thank you for your question. I mean, it is true that the AI development is happening quicker than anyone can react to. And we see it around the world, frankly. And also we heard this from representatives of scientific panel today. So the question really is, should we let it go then, or we do something? And I think we've been trusted with doing something meaningful, and because my dear co-chair explained how we came to the dialogue,— and we came to establishment of also scientific panel. So I think that the next dialogue will happen in New York. The scientific panel is entrusted to do the work for 3 years, so I guess that we will produce many other interesting and useful reports and so on. So we were able here to start something, meaningful discussion, on a global level, create a platform, and to address few issues. There are 7 prescribed topics in the modalities resolution that we compressed into 4 clusters, and there will be 4 thematic discussions during these 2 days. This is one thing. But secondly, the global landscape is really fragmented, and there are pockets of knowledge here and there on a regional level, in countries, and so on. So Our goal, so to say, is to be maybe able to reduce this fragmentation and at the same time try to build interoperability between different elements, whatever. And for this, the timing is actually perfect, I would say. And I think that also the ITU Secretary-General said that, you know, it took decades to deal with internet, but with agentic AI, we are lagging behind 3 years only. But I think so we have to— the main point is that we have to start this dialogue somewhere. And today we did it. And this is, of course, indeed coming from GDC and our established resolutions of the General Assembly. Thank you. I'll take a last question.

Nina?

Thank you. Nina Larson with AFP. So just, I wanted to be sure, is there anything concrete that you're expecting at the end of this? Will there be a joint statement or anything? And also, when it comes to the Child Safety Pledge that was mentioned, is that ready to be endorsed yet? Where, where do you see that going? And finally, just one last thing, more generally, I'd ask you, what do you think is the most frightening thing, I guess, because you've been painting, I guess, some frightening pictures also of what AI is capable of. What is it that keeps you up at night, and how confident are you that you're actually— that this UN-driven process will actually be able to harness this? Thank you.

I will leave the last part to you.

3 questions. Yeah, please go ahead.

Okay, thank you. I guess that concrete outcomes My co-chair was already mentioning the co-chair summary, so he also mentioned about the written submissions that we received from a variety, not only from member states but stakeholders, that conversations that we have hearing throughout the process plus today and tomorrow will be part of what we're going to compress in a co-chair summary. Going back to the modalities resolution, It has already been established that this, together with the report of the scientific panel, will serve as the foundations for the conversation that member states will have for the second dialogue in 2027, and also the basis for common understandings on AI governance that will be decided by member states next year. But not only that, as part also of the process, we have established the AI Partnership Hub. This is a voluntary platform that member states and different stakeholders have submitted initiatives that are already happening so we can encourage more interoperability but also further cooperation. And this is something that we really hope to see throughout our consultations. We were and from many stakeholders, but also member states, we want this to be more than simple speeches. We want something concrete. So as I said, we're setting the foundations, but we also wanted to provide something practical that can encourage the cooperation, conversation, and also more action-oriented results. And that's why the AI Hub platform is one of those. PJA?

Yeah.

What keeps me, I think all of us, awake at night is that we are not quick enough to protect our children. And I think this is really the dimension where especially those who are providing these platforms, who are feeding the algorithms who are designing this new world have to not only see their role, but their responsibility. Because in the end, this will come back to all our own homes and all our own children's bedrooms. And the examples we have heard this morning, also on stage, this is happening right now. And we have laws all over the world to protect the dignity of human beings. And the core foundation is the Charter of United Nations, with all its three principles on peace and security, sustainable development, and human rights. And we have heard that very often in the United Nations, with the argument of being neutral. And we hear it here also very often as well. About disinformation, that how can you take a side? But I think the important thing is to always remember that inaction also means taking a side. So if there is a spread of misinformation used as warfare— and I was quoting this morning a teenager from Rwanda, a country which have experienced genocide, And if a teenager is warning these days in a commemorative meeting about prevention of genocide, how these social media platforms can be misused to prepare other genocides in the future, this is the responsibility of not hiding behind so-called neutrality, but putting safeguards in the social media world of preventing genocide as well. And again, this is not about taking a side. This is— if everybody has signed a Genocide Convention— implementing a Genocide Convention in the social sphere. This is about implementing the Universal Declaration of Human Rights in the social sphere as well. If we see that most of the attacks with sexual deepfakes are targeting women and girls. And there we can also see the power of societies and the power of markets. We have seen, for example, in India that their answer to one of these apps, Grok, which made possible that you have a photo of women, of girls, and they are being undressed in the internet. If If a member state, if a country, if a society says, "We are not allowing this against our human rights, our women's and girls' rights, otherwise this app cannot function in our markets," obviously, within a very short time, this feature was disabled. And this underlines again that we don't have to wait until the next Global Dialogue in one year's time. We don't have to wait to set up a global governance structure that all 193 member states agree on in common. We have seen over the decades of the United Nations that it always needed a cross-regional alliance to push things forward. We have seen that in the climate fields. We have also seen it in the international human rights fields, that if member states move forward together, They can frame and shape our future and the benefits of all. And this is also what this global dialogue is about. The co-chair was underlining the steps which the General Assembly is taking together, but this doesn't hinder companies and governments, either together or by themselves, to move forward as frontrunners. And we have seen different member states all around the globe which are introducing now already AI governance structures.— and therefore also the learning from each other, being better together by joining forces. This is one of the key parts of these days here in Geneva, and therefore this summit, the Global Dialogue itself, is already a big step forward in making the future for children altogether safer. Thank you.

Thank you very much. Would you like to add something? Then we will close.

Yeah, no, thank you. Thank you very much. As I said, I'm father of two beautiful daughters, and of course I care about their future. And, and so the one thing that keeps me awake at night, of course, is how we are going to protect our children from bias, from intimidation, from sexual abuse that we see these days being conducted using AI. But I also wear another hat because for this year I'm President of Executive Board of UNICEF, organization that was brought to life 80 years ago exactly to do this, to protect children. And I think for these 80 years they've done a great job. But now we have to together somehow to design another I don't know, element of this work which is linked to exactly to AI. And so we've tried to provide as much as possible support to UNICEF, you know, to build structures that are necessary to do it. And of course, member states are very, very valuable. There are different coalitions of the willing. We've seen it in the past. We see it now. You mentioned one example. I think that protection of children is something where we can put aside maybe our ideological and political, you know, differences and really to be united in delivering to protect our children. So I think that this is a great initiative, and of course I hope that it will be built also will bring tangible results. The second element, let me just mention, is that, you know, when we first met with Secretary-General, he mentioned that on certain point not so long time ago, there were more than 100 countries that had no involvement whatsoever in AI affairs. So It's a huge number, and I guess that this is something that really motivated us as co-chairs with this dialogue to bring as many countries and stakeholders together—civil society, academia, think tanks, industry—and to conduct this first dialogue. And today we heard that there are 107 of the 170 countries represented here. And just to remind that there is 193 countries in the United Nations organization. So we are not yet there, but I think it's much more than it was a few years ago when the Global Digital Compact was established. So we are moving forward. We are inclusive, trying to bring these different perspectives and voices here, but it will require much more. That's why I hope that with a second dialogue in '27, with a review of a global digital compact, we will be able actually to really to bring also these remaining countries and to have a meaningful discussion, all of us together. Because this, I mean, AI impacts everyone, whether you like it or not, whether you are present or not. We have to do something with this. I think the dialogue is the best thing That would happen, and indeed, thank you very much for quoting me. I hope that this would be a San Francisco moment for AI, and we will perceive it 10 years later as a defining moment when we started this discussion.

Thank you. Thank you very much. It will be the dialogue of Geneva. With that, I'm afraid we have to close. Thanks for being here, and please enjoy the rest of the dialogue, and bon appétit.