Foster AI accessibility for building inclusive knowledge Societies: a multi-stakeholder reflection on WSIS+20 review

UNESCO Session Explores AI Accessibility for Inclusive Knowledge Societies

The UNESCO session titled “Fostering AI Accessibility for Building Inclusive Knowledge Societies” was a comprehensive discussion on the role of artificial intelligence (AI) in promoting digital inclusion and the integration of marginalized groups. Moderated by Xianhong Hu, a program specialist from the Information for All Programme (IFAP), the session featured a panel of experts who provided their perspectives on the impact of AI on society, particularly focusing on the concept of meaningful connectivity and the inclusion of diverse voices in the AI value chain.

Onika Makwakwa, a founding partner of the dynamic coalition on measuring digital inclusion, emphasized the transformative potential of AI across various industries. However, she warned against the risks of excluding marginalized groups from the development of AI systems, which could lead to biases and further societal divides. She stressed the importance of inclusive digital development, advocating for meaningful connectivity that encompasses quality access, digital literacy, and affordability.

Alexandre Barbosa, head of CETIC in Brazil, discussed the shift from binary measures of connectivity to a more nuanced understanding of digital inclusion. He highlighted the Brazilian Observatory on AI, a collaborative initiative aimed at consolidating information on AI deployment in Brazil and engaging various stakeholders. Barbosa also mentioned Brazil’s G20 presidency and its focus on AI, meaningful connectivity, misinformation, and digital government.

Fabio Senne, also from CETIC, raised concerns about inequalities within AI, such as the representation of vulnerable populations in training data and the impact of AI literacy on the population’s ability to mitigate technology-related risks. He pointed out the need to address the new gap between those who have access to digital technologies and those who are empowered to make informed choices.

The session also saw the announcement of the Internet Governance Forum Dynamic Coalition on Measuring Digital Inclusion, which aims to develop comprehensive metrics for digital inclusion. This coalition is a step forward in promoting policies for inclusive, equitable, and sustainable knowledge societies.

Maja Maricevic, Director of Science and Innovation at the British Library, spoke about the evolving concept of digital inclusion in the AI era. She highlighted the need for actions that address complex issues such as digital rights, democratic participation, and ethical safety. Maricevic suggested practical steps like demystifying AI by educating the public about its various technologies and implications, promoting human-centric design, and creating open environments for experimentation with data and AI tools.

The session concluded with a call for ongoing dialogue and multi-stakeholder partnerships to ensure AI’s role as an empowering and equitable tool. Xianhong Hu called for collaborative research and policy recommendations to address the new digital divide in the AI age. Participants were invited to join the dynamic coalition and contribute to advancing digital inclusion metrics.

Key observations from the session included the recognition of the innovative potential of diverse teams in AI development, the importance of understanding AI’s economic models, and the emphasis on AI literacy as a foundational element for inclusive digital development. The need for a multi-stakeholder approach to AI governance, learning from internet governance structures, and ensuring civil society participation in AI governance discussions were also highlighted.

Xianhong Hu:
Thank you. Colleagues, if you allow me, let’s have another minute to have our chair to get his water. I think we can wait for one more minute, but I could just start to give some background of this session. Basically, welcome you all to this UNESCO session on Fostering AI Accessibility for Building Inclusive Knowledge Sciences, organized by the Information for All Program, jointly with our working groups on the information accessibility and information ethics. I think we also have some working group members joining us online and perhaps also in the room. Thanks to the digital technology, it’s my first time to moderate to the session remotely and actually I find it quite functional. My name is Xianhong, the program specialist in the Secretariat of Information for All Program. We are today having a tight schedule. We are having a fantastic panel of five speakers, each of them are supposed to give five minutes intervention to share their visions and views. I also like to have some time really reserved for the questions and answers in the end. But since it’s a hybrid session, for those who already have some comments, suggestions or questions, you can just type your comments in the chat. I also take this opportunity to introduce my team, our colleague, Mr. Yacoub Dutoit, also joined us from the HQ, and Victoria Gruduc, and also Ludovica. We are all joining the session, they’re supporting the discussion from Paris. I don’t feel there’s such a distance, basically, we can see the room and here. everything clearly. So we are more or less on the time. If we should wait for a slightly a bit more to have the chair to address the welcome remarks because that’s quite important for us to set a scene on today’s discussion. But if we still need to wait for some time perhaps let me take a maybe let’s take a bit of action forward. Today we are having two objectives. One is to trigger discussion and raise awareness on how we ensure the inclusion, ensure the participation of those marginalized groups into the development and evaluation of artificial intelligence. And we also have a second layer of objective to catalyze a partnership with the IFA working groups and also through our newly launched dynamic coalition on measuring the digital inclusion. I happen to have our first speaker actually Ms. Onika Makwakwa who is also a founding partner of the establishment of the dynamic coalition on measuring the digital inclusion. While we are waiting for the chair I would like to give the floor to Onika Makwakwa since we are the founding partner on this time. Maybe you can start to share some view on how you think about the impact of artificial intelligence and also other frontier technology. How does it impact the issue of digital inclusion? And what do you think that maybe your organization, DDIP, could contribute to support the further action on the digital inclusion, as well as through this dynamic coalition? If you don’t mind, Onika. So we are having the chair coming. So if you don’t mind now, I would like to give the floor to the chair before we start the panel. But more or less, I think you have already heard of what the session’s about. Okay, so hi, Mr. Chair. Hello. Good morning.

Pablo Medina Jimenez:
Thank you very much, Ms. Cheng Hong. A pleasure to have you online.

Xianhong Hu:
Yeah, Mr. Chair, I just to give you a very short briefing. Before you arrived, I just briefly introduced the objective of this workshop and also briefly about the scenario. So now I’m very honored to introduce Mr. Pablo Medina Jiménez, the chair of the Information Forum Program to give some welcome remarks. And also I’d like to thank you for your taking efforts to be present in Geneva. It’s your first visit to the forum and I wish you really enjoy it. So Mr. Chair, floor is yours.

Pablo Medina Jimenez:
Thank you very much for your kind words and for the introductory remarks. As I was saying, it’s a pleasure to be here among all these distinguished representative colleagues, experts. So I will start a very brief presentation, if you allow me, Ms. Cheng Hong. Distinguished participants, esteemed colleagues, it’s a privilege to welcome you today to this important gathering at the WSIS 20 review. We are here to engage in a pivotal thematic workshop titled Fostering AI Accessibility for Building Inclusive Knowledge Societies. organized by UNESCO’s information for our program. Reflecting on the past two decades, we observe a digital transformation that has reshaped every aspect of our society. As we navigate this era, frontier technologies like artificial intelligence, big data, have the potential to either bridge or widen existence societal gaps. So it’s our duty to ensure that the development of these technologies benefit everyone, ensuring no one is left behind. Information for our program, established in 2001, is a unique UNESCO intergovernmental program that champions equitable societies by promoting universal access to information and knowledge for sustainable development. Our mission revolves around six key areas. The first one, information for development, recognizing the value of information in addressing development challenges. Two, information literacy, empowering individuals to seek, evaluate, use, and create information. Three, information preservation, ensuring universal access by preserving libraries, archives, and museums. Four, information ethics, addressing ethical, legal, and societal aspects of ICT applications. Then information accessibility, focusing on availability, affordability, and accessibility of information. And the last one, the sixth one, multilingualism in cyberspace, which facilitates participation in knowledge societies. IFAP is committed to promote information accessibility for all and to promote an ethical approach to new technologies. Our mission, as outlined in the IFAP strategic plan 2023-2029, focuses on enhancing information accessibility, information literacy, information preservation, information ethics. information for development and multilingualism in cyberspace. These pillars are essential in building sustainable and inclusive knowledge societies. Strong partnerships also are fundamental to realizing our goals. I am pleased then to announce today the establishment of the Internet Governance Forum Dynamic Coalition on Measuring Digital Inclusion. This new dynamic coalition created in collaboration with the Global Digital Inclusion Partnership and key IFAP partners, including the International Federation of Library Associations and Institutions, the Regional Center for Studies on the Development on Information Society, CETIC, the Governance and the Tech and Global Affairs Innovation Hub, will lead efforts to develop comprehensive metrics for digital inclusion. At the heart of the coalition’s mission lies a commitment to mainstream policies for inclusive, equitable and sustainable knowledge societies, international development plans and digital transformation process. The coalition also considers addressing gender divides in digital access, literacy and participation, and fostering opportunities for women’s participation and leadership in the digital age. We invite all stakeholders, experts, governments, representatives, private sectors, leaders and members of the academic and technical community to join this coalition to contribute to its advancement. In conclusion, dear colleagues, today’s workshop offers a chance to start discussions, build partnerships and gather resources for inclusive innovations. Let’s use the collective strengths of our networks to ensure that AI and other emerging technologies serve as tools for empowerment and equity. Thank you very much for your participation, your commitment to building inclusive knowledge societies. And let’s continue this work beyond today’s session. Thank you very much for your attention.

Xianhong Hu:
Thank you. Thank you very much, Mr. Chair. for setting such a good scene for our following discussion. I also take this opportunity to inform that we also have the presence of the former IFEP chair, Madam Dorothy Gordon online. It shows such a strong support from the previous IFEP leadership. I’m pretty sure today we’re having a very inclusive discussion on this important subject. Onica, now you have the floor. As I just mentioned, I’d really like to have you to share your vision and the view about how AI is impacting the digital inclusion issue and how you and your organization, Global Digital Inclusion Partnership would be able to action on it. Thank you. Wonderful.

Onica Makwakwa:
Thank you so much for this opportunity and good morning, everyone. You know, the Global Digital Inclusion Partnership is steadfast focused on advancing meaningful connectivity for the global majority. With over 10 years of our dedicated work on digital inclusion through our team from affordable access to meaningful connectivity, we recognize the imperative of ensuring inclusion and participation of groups left behind throughout the AI value chain. AI holds immense promise to revolutionize industries, to improve efficiencies and enhance our lives in many countless ways. However exciting this seems, it’s important that we don’t overlook the importance of inclusive digital development as a foundation for every stage of AI development. Imagine a world where the voice and perspectives of women, rural communities, racial and ethnic minorities, persons with disabilities and low income communities. and 1.5 billion people who are unconnected, mostly from the global maturity world, are absent from the creation of AI systems. Unfortunately, this scenario is not too far fetched for us, as marginalized groups already face barriers that limit their participation in the AI value chain. Why does this matter, we would wonder, right? It matters because exclusion breeds bias. When AI development teams lack diversity, they unintentionally embed their biases into algorithms, perpetuating inequalities and exacerbating societal divides. From biased hiring practices to discriminatory predictive policies, the consequences of exclusion in AI development are far-reaching and quite profound. But it doesn’t have to be this way, right? Inclusion and participation are not just moral imperatives, they are prerequisites for building AI systems that are fair, accurate, and beneficial for everyone. Research has actually shown that diverse teams lead to more innovative solutions. And this brings me to one of the things that is really critical in this foundation of AI, and that is making sure that we have inclusive digital development, where everyone is meaningfully connected and able to participate in a way that can truly be transformative for their lives. One of the big questions is, what could be the impact of AI and other technologies on digital inclusion? We would not even be able to honestly get the full benefit of AI if it is only developed by just a select few and those who are connected at the moment. So I think the imperative and the… opportunity now is that because of this interest and growth of AI, it actually provides an imperative for us to truly double down and fast track our goals for digital inclusion. And in doing so, it’s also important that we look beyond basic connectivity. When you look at all the things that AI promises to do and can enable us to do, it becomes an imperative for us to look at how we are measuring access and how we are measuring connectivity, like what is a connected person, and really go back to some of the work we had started around meaningful connectivity to raise that work so that we raise the bar for the international standard for what a connected person is and what it means to be connected in a way that also can enable you to contribute towards and or benefit from the artificial intelligence ecosystem that we have at the moment. So we have a lot of work ahead of us to make sure that policies for digital inclusion are not only just about connecting people, but making sure that from all the lessons that we’ve learned, including through the COVID-19 lockdowns, that we are now raising the bar to make sure that we are also closing the emerging gap amongst those who are connected. And that is looking at the quality of the connectivity, looking at the utility in terms of them being able to have the requisite skills and the affordable devices to be able to fully enjoy the opportunity of being connected, and just literally the entire spectrum around meaningfully connected as opposed to just basically being connected. Thank you.

Xianhong Hu:
Thank you so much, Onika. Just to reinforce what you have said about the role of advocating measurement. on digital inclusions through our digital dynamic coalition. We had just launched the membership submission portal. So we are here really, we joined to invite all the stakeholders online in the room and at WSIS to join us on this joint endeavor. You only need to submit a very short form on what you are going to do to contribute. So we will be able to collaborate together and engage you further. And now I’d like to introduce the second speaker, Mr. Yves Poulet, the Emeritus Professor and Director of the University of Namur in Belgium. And he used to be the chair of IFA working group on AI ethics. He’s truly an expert in the regulation and policy of AI. So Professor Poulet, would you be able to take the floor? Hello Yves, are you able to hear me? Hi Yves, we couldn’t hear you clearly. Could you try again? We still cannot hear you. Yves, I think there’s a connection problem. Basically your voice are broken from time to time. No, we still cannot hear you. Would you be able to maybe… exit to try again, and my colleague Victoria will be able to help you. If you don’t mind, I’d like to have our next speaker to talk until the technical problem on your side will be solved. Is that okay with you, Eve? Okay, good. So now I’d like to introduce our two speakers from our long-term partner, CITIC, which means the Category 2 Institute of UNESCO, the Regional Center for Studies on the Development of Infectious Diseases, based in Brazil, the head of the center, Mr. Alexandre Barbosa, and also the expert, Fabio Sanna. I mean, thank you for attending our session and also sitting on the panel in the room. So, Alexandre, would you like to take the floor first?

Alexandre Barbosa:
Yes, thank you very much, Xianhong, and thank you, Mr. Chair, Pablo, for inviting us to be here. Good morning, everyone. It is indeed a great pleasure to be here, as Xianhong has mentioned. We have a partnership with UNESCO for a long period now. And in the context of this new dynamic collision on measuring digital inclusion, I think that, Xianhong, it is indeed a very important initiative because it does touch several WSIS action lines. And in my opinion, it is a very relevant initiative from UNESCO, aiming at providing insights on advancing beyond the simplistic binary measurement of connectivities, as we are used to discuss, to have or not having access, and to move to a position that advocates for a more complete comprehensive approach to digital inclusion and the debate and research around digital inclusion is one of our priority areas at CETIC, the Regional Center for Studies on the Development of the Information Society at the Brazilian Network Information Center. We are a data producer center in Brazil, and since 2005, we have been conducting surveys, research, capacity building, producing indicators for policymaking on the socioeconomic implications of digital technologies. So this is indeed a very important area for us. And as a UNESCO Category 2 center, we also cooperate with Latin America and the Caribbean countries and Portuguese-speaking countries in Africa to develop better and internationally comparable data related to digital inclusion, and of course, the implications and the data analysis around this topic. And in terms of the intersection of digital inclusion and the adoption of artificial intelligence, we have been working in the past five years or so on new indicators related to AI adoption in different areas of society, especially by enterprises, by government, how government are adopting AI to provide better services or to provide a new way to interact with the citizens, and also the adoption of AI in the health sector and also education, how those establishments have been using AI to provide better services. And particularly in terms of AI adoption, we have been also conducting… not only the traditional representative sample surveys that produce reliable data and representative data, we have also been conducting sectoral studies on AI. We have launched two years ago during the UNESCO World Conference on Culture, a publication on the implications of AI in the cultural sector. And I have the pleasure to announce that in the second semester of this year, maybe September, we are going to launch our publication on AI in the health sector. This is a quite important study. And in this regard of AI, I would like to mention that we are partnering with the Minister of Science and Technology and Innovation to implement the Brazilian Observatory on AI. And we are responsible for making this happen. It’s not only ourselves, but we have four key partners, including the Brazilian government. And this initiative will bring together different stakeholders and consolidate information from different sources in terms of deployment of AI in Brazil. Well, now I would like to give the floor to Fabio because it is very important to understand how AI relates to potential inequalities in different areas, such as how AI can be biased in terms of the data, the models used to be trained, or the tools or the lack of skills. And last but not least, I would like to mention that recently, one month ago, we published, under the G20 Digital Economy Working Group, the publication on meaningful connectivity. And Brazil has embraced this area, because we have to go beyond the binary states of being connected or not, to include, for instance, digital literacy, digital skills. So I would like to give the floor to Fabio to explore, Fabio, maybe, these issues of inequality and AI in terms of digital inclusion.

Fabio Senne:
Thank you, Alexandre. Thank you, Mr. Chair. And thank you, Shanhong and IFAP, for the invitation. Yes, I would like to comment a little bit on the evidence we have in Brazil, for these 20 years, on measuring issues related to digital inclusion, and now on the discussion and interaction with AI. I think I’d like to raise three points that I think are critical to understanding this discussion. I think the first of them, I think, Ms. Onika already mentioned, which is how to deal with the very beginning of AI, or the design and development of AI systems and models, and equality of the training data. So if you don’t have representative data from different populations and vulnerable populations, of course, you’ll be biased in the way that you train your AI models. So I think it’s important to raise the questions, not just from a perspective of inequalities within countries, because we know that the levels of digital inclusion vary a lot between countries and regions of the world. So how can we ensure that the most vulnerable, in terms of digital inclusion, are represented in AI models, but also the inequalities within countries that we know that are very large. This is also the case of Brazil when you compare, for instance, ethnicity, traditional populations, rural versus urban, income, level of education. So when you go to all these dimensions, you can see that those inequalities will also be represented in the AI models that are being developed. I think that there’s another second point on the use of, because now we know that with the large language models and generative AI, we know that every one of us can use also, can be very active in using AI-based solutions. So in the use perspective, also the inequalities will have a very important impact. So when you go, for instance, we know from other technologies that those that are early adopters tend to benefit more in a faster way. And those that do not have the experience to use those types of technologies tend to be left behind. And this is a way of generating new inequalities and increasing inequalities if we don’t take care of the use dimension. And finally, I think this is something that Omnika also mentioned, is the skills part of the generation of AI literacy among the whole population. We know from the evidence we have that there is also the most vulnerable population in terms of digital inclusion are also more vulnerable in facing the risks related to the technology. And this can have to do, for instance, with misinformation or disinformation process that we face in societies. And from the surveys we have, we have a survey, a very interesting survey. children in Brazil, which is more following the methodologies of Kids Online Survey, which is defined by Europe and UNICEF. And we know from the surveys that although children are very well connected in the country and using AI tools in a very innovative way, we know that, for instance, 40% of Brazilian children from 11 to 17 agree that the first result found online is the best result. 51% agree that every person finds the same content when searching online. And 42% are unsure about their own abilities and skills to check online information. So when we ask children about this, we know that from the very beginning, we need to develop online and digital skills that can make this type of AI implementation useful and inclusive and not create new types of inequalities. I would like to just raise these three points that I think are critical for our discussion and also to welcome this new IGF dynamic coalition that I think this can be a very useful space for developing this debate. Thank you very much.

Xianhong Hu:
Thank you so much, Alexandre and Fabio, for sharing so much substantial and pioneering work relating to meaningful connectivity and unpacking the complexity regarding the digital inequalities of AI chain and development, which is so inspiring. Please do share all your work with us so we can further disseminate through the IFAP network. I already copied the link of your new report on meaningful connectivity online. It’s a really wonderful report for everyone to have a check. So now let me try if Professor Yif Pulei, would you be able to try again if you can take the floor? Yif, are you there? Hello, perhaps we need more time for Professor Poulet to intervene. So now I’m lucky to introduce Dr. Maja Maricevic, the Director of Science and Innovation of the British Library. As you know, the library is playing an instrumental role in meaningful connectivity and fostering the digital inclusion, particularly in the AI age. And the British Library has done so much wonderful work. So Maja, please take the floor.

Maja Maricevic:
Thank you so much, Xinhong, and thank you so much to all the previous speakers. It was really good to hear the emerging consensus about the importance of the issue that’s facing us in this particular area. We’re so excited to be here and especially about the dynamic coalition, because I think the coalition of different stakeholders is exactly what’s needed in this particular moment. So we had an excellent session yesterday with also colleagues from IFLA talking about digital inclusion, and I spoke at that. And in some ways, I’m going to continue from where we stopped yesterday. So we’re talking about digital inclusion and libraries, and I was trying to suggest to colleagues, and I think I heard some of these views from other speakers today, that there is a very big change in how we’re seeing digital inclusion. And before AI, we simply saw it as closing the gap before those who do not have access to digital technologies and those who do not. And this issue of access obviously stays with us because there is still very large parts of communities internationally that do not have any access at all. But with AI, the situation obviously has become lots more complicated because now we have a new gap that I think everybody was referring to. And Fabio just gave us some amazing statistics from Brazil that we also now have a gap between. those who don’t have choice, how they access digital technologies. They might have access, but it’s very narrow and they are not empowered to make the right digital choices for a number of different reasons. And there is no doubt that AI offers us new economic opportunities and some amazing advancements, opportunities for the humanity, but also causes adverse effects by AI and data-driven technologies and obviously algorithmic profiling that leads to harms, bias and misinformation that’s already been mentioned. And also there is lots of issues, I think, in the area of digital rights that I’ll come back to in a little while, a little bit more. So for libraries and people who work with communities in different information settings, obviously the actions that we have been taking, which is about lending devices, making sure there is access to Wi-Fi, providing digital literacy education and many, many other things. There is now the whole additional consideration of different actions that are now related to very complex issues. I mentioned digital rights, including digital cultural rights, because internet does not have the equal amount of data information from different cultures and societies. There is the whole level of issues about democratic participation in digital decision-making. There is lots to be said about how we promote human-centric design to different systems and services, but there’s lots to be done on domestified technology and what it actually means, and to also take greater care of diversity of digital outputs in terms of linguistics, cultural, geographic, economic. Think how we. preserve digital information with AI in it and how we empower parents, teachers, citizens, creators, employers to make right informed digital choices and then the big issues around ethics and ethical safety issues. I want to just quickly do that very big list that just as we were talking about accessibility that maybe there are some practical steps that we could be thinking about. And I’ll suggest a few and see if people agree and if our discussion maybe leads to something there. One for me that is big on here is the mystification of AI. So we talk about AI all the time as one single technology but in effect we are dealing with many different technologies. They all have their advantages and disadvantages. We have machine automation, robotics, machine learning, large language models, computer vision, neural networks, theory of mind. There are many many of them. They all have different implications. They’re used in different ways. They have different advantages and different limitations and I think there is something about seriously upping education levels of people understanding what these technologies are and what they actually do and where they appear. So with large language models it should not be that difficult for start understanding that they simply use statistical models to analyse language and the different things can happen as a result of that and so on. So I think one thing I’m suggesting is that we need to get under the bonnet of AI and maybe not talk about it as this magical one thing because I don’t think it is a magical one thing. The other element for me is human element. So we talk about technology quite a lot but the human element and the human agency in AI is another thing that I think really needs to be separated out because we have human agency in designing and employing. employing AI. Legal and ethical aspects are really up to humans to make sure they’re responsible for, and humans have choice whether or how to use AI. As well, humans have rights to question AI-derived outcomes because AI is often very, very wrong. And I think that’s something that we really have to have on our agenda. And also make sure that we put in our education, understanding of AI economic models, because AI is there to give us a really new scientific breakthroughs, resolve health issues, make big human advancement, but it’s also there to make money. And all of these things are good things. Economic advancement is a good thing. But I think we have not maybe put enough stress on working with people to understand AI economic models, and also how all of us participate in this as data points in our own right. Our voices, our data that we leave on the internet, everything that we are becomes a part of this economic model. And I think working with people to understand that better would certainly, I think, increase understanding and inclusivity of this. And my last point is about participation and experimentation. So the libraries do quite a lot in making our collections available and our data available to people in accessible ways with different tools. Obviously, not everybody has the same resources, and we are very fortunate to have resources to do this. But I think there is something about making sure that across the world, there are open environments where people can play with data, play with tools, and learn by doing. And this can be made very accessible for digital storytelling, through crowdsourcing exercise, from free tools that are provided in a more… more responsible and ethical environments and I think that’s something that would be really important to do because then this technology will start to feel real and we will have more understanding how it actually works rather than being something that is just coming from somewhere else. So I’m so excited about the idea of coalition of different partners coming together on this because I think that combining of skills of different agencies and parts of the academia and the governmental organization and the international organization would be exactly the right way to go about this. Thank you.

Xianhong Hu:
Thank you Maria for sharing so many brilliant ideas and also sharing a great work talking about the humanistic aspect for AI. I think no one can really talk better maybe than our next speaker and Professor Puley. I think I just saw you. Could you try again? Hi Yves. Hello. Hi Professor. Could you try to turn on your camera and present a bit? Actually I saw a double screen of Professor Yves Poullet. Somehow it just showed up but now it didn’t. So Professor Eve, I hope you can hear me. So whenever you think you could talk. Ah yes. Hi Eve. I saw you. Can you hear me? Hello? Hi Eve. Oh you cannot hear me. Okay. Anyway I hope that you can speak whenever you can. So now we only have a few minutes left. I’d like to open the floor to the participants. in the room and online. I also like to ask Mr. Chair, Mr. Pablo Medinista, could you please help to co-moderate a bit if any participant in the room who want to take the question to share some comments?

Pablo Medina Jimenez:
Yes, team Xianhong, of course. Thank you very much for considering this possibility. So if you allowed me, as you are the moderating one, to ask here in the room if there is anybody who would like to take the floor to any comment or question, anything, feel free of intervening in this workshop. I see. Could you please present yourself, introduce yourself, and take the floor? Thank you.

Audience:
Thank you. I’m Amanda Leal. I work at the Future Society. We’re a nonprofit focused on AI governance. And I came here because I was interested also as a Brazilian to hear more about the work that CGI does and understand what we can learn in AI governance from internet governance and these structures that have been in place for many years. And one thing that I came here, I came here expecting that I would make an intervention arguing that it’s not only about accessibility. And luckily, I heard from all of you that it’s not only about accessibility, so this is great. So taking a step forward, because I think we’re all on the same page about the need to address AI economic models, power concentration, digital inequalities, media, and information literacy on top of digital literacy, I wanted to know from you in a positive agenda, how do you envision critical things that have happened from internet governance? What should we learn and take forward in AI governance? And I will make sure to try to reflect that as well as a civil society representative in this area. it’s more and more crowded with industry lobbying and interests that are not necessarily aligned with public interest because the economic incentives are not there. So what would you suggest in terms of bringing this bottom approach and having more civil society participation in AI governance? Thank you.

Pablo Medina Jimenez:
Thank you very much, Ms. Leal, for your remarks. I see Mr. Alexander, would you like to go to the floor? Okay, yeah.

Alexandre Barbosa:
Thank you so very much, Amanda, for this question. Indeed, a very critical question. As you may know, Brazil holds now the G20 presidency and it is not a coincidence that our four priority areas within the digital economy, a working group, we have AI, meaningful connectivity, misinformation or information integrity, and digital government. And we are involved in these four areas, but particularly in providing guidelines in terms of AI and meaningful connectivity. We have two knowledge partners being UNESCO, our knowledge partner for AI and ethics, and ITU on meaningful connectivity. And of course, you brought a very important message on how to engage different actors from society. And as you probably know, the Brazilian Internet Steering Committee, CGI.br, since its inception, a multi-stakeholder structure, nature. So we bring to the table different actors from society. And right now, the topic of disinformation related… to AI and information integrity as well as the regulation of AI, we have to, this is an ongoing process right now in Brazil and we are listening to different stakeholders so that we can ensure that we can incorporate different voices into this debate. Yesterday we saw an excellent presentation from UNESCO in terms of AI in the judiciary. We have been also working with UNESCO, the first MOOC that UNESCO has developed including your own organization, the Future Society and ourselves and of course UNESCO. We are very much concerned and worried with this potential bias that we have in terms of using AI-based applications in the judiciary. So this is not a completed discussion but it’s an ongoing process right now. But thank you for bringing this important issue. Thank you.

Xianhong Hu:
Mr. Chair, may I also intervene on this a bit? I mean today we are really focusing on a very crucial issue on the meaningful connectivity and inclusion regarding to the artificial intelligence which is a new cross-tech area. IFAP working group on information accessibility has been convening a milestone annual conference on artificial information accessibility related to the artificial intelligence. It’s called AI for AI. IA for AI means information accessibility for AI for information accessibility which will be convened every year on 28th September also in the commemoration of of the International Day of Universal Access to Information. I think that will be an important occasion for us to really carry forward what we’re discussing here. I think we do need to consider conducting more collaborative research to come up with some policy research, a brief recommendation to really sensitize our member states to tackle this emerging new issues on the inclusion regarding to the bias, inequality, et cetera, relating to the algorithm, data, infrastructure and application of the AI development. It’s become a new emerging digital divide in the age of AI. Age of AI before we have not succeeded in tackling the digital divide emerged by the internet, et cetera. So it’s definitely a new area we should tackle. And I welcome you all to join our IFAP working group for accessibilities activity in September. It’s called the AI for AI. I will also type a link to the chat. And also as a dynamic coalition, as all of your speakers have well supported, I mean, you are all our actor and driving force. We are going to launch this dynamic coalition with different stakeholders on different occasions throughout the year. First coming one will be at the EuroDIG. For those who will be there, we can further strategize the actions. And then we are going to have the iSCOV conference, which will take place in South Africa later this year. And then in December, we will have this dynamic coalition gathering in the ITF to be convened in Saudi Arabia. So let’s not stop the conversation on this, but really continue enlarge and deepen it because it definitely is one of the most crucial issues for IFAP, for our stakeholder to tackle in the future. And I think we’re almost at the edge of the finishing our session. I also like to. to have a final check for any speakers, panelists in the room and online. You have any final remarks to make? I saw Maja send a very nice smile and I take as a very good message. And also, if not, I’d like to give floor to Mr. Chair, Pablo Medines. Maybe you can say really last word to close our session. I thank you again for being present and also hybridly moderate a session with me. And Mr. Chair, the last words is for you.

Pablo Medina Jimenez:
Well, thank you very much, dear Xianghong. I didn’t prepare anything for closing remarks. Indeed, as you’re the one who is in control, I wanted to ask you the permission to ask the floor, to ask the room if there’s any other reactions about our interventions we had here. So if I may, then thank you and everybody for convening this excellent workshop. And of course, a salute to Madam Dorothy Gordon, which I didn’t know when I started. She was online. So dear Xianghong, I give you back the floor for your closing considerations as you’re the moderator in this workshop. So thanks again. Thanks, everybody, for having me here with you. And Xianghong, the floor is back to you.

Xianhong Hu:
Thank you. Thank you. Thank you, Mr. Chair. Thank you, everyone in the room and also online. I just want to last try for Professor Poullet can join us. It seems impossible, but I do. Yeah, and I apologize because we have the next. session coming. Yeah so probably next time. Okay and thank you. I also like to thank the excellent technical support of WSIS Forum. Thank you madam. I really enjoyed the whole hybrid session which went so smoothly. Thanks to all. Have a good day and enjoy your WSIS in the forthcoming days. Bye. Thank you. Bye-bye. Thank you. Bye.