Global Digital Governance & Multistakeholder Cooperation for WSIS+20
10 Jul 2025 10:00h - 10:50h
Global Digital Governance & Multistakeholder Cooperation for WSIS+20
Session at a glance
Summary
This discussion focused on strengthening inclusive, rights-based digital governance as part of the World Summit on Information Society (WSIS) Plus 20 review process, with particular emphasis on artificial intelligence ethics and information integrity. The session brought together representatives from various organizations including the ITU, European Commission, UN Office of the High Commissioner for Human Rights, Wikimedia Foundation, and Internet Society to explore multi-stakeholder approaches to AI governance.
Rasmus Lumi from Estonia’s Ministry of Foreign Affairs, representing the Freedom Online Coalition, opened by emphasizing the critical importance of maintaining the multi-stakeholder Internet governance model against attempts to impose centralized state control. The panelists consistently stressed that effective AI governance requires meaningful participation from all stakeholders – governments, private sector, civil society, academia, and end users – rather than relying solely on multilateral approaches.
Key themes emerged around the need for transparency, accountability, and inclusion in AI systems. Isabel Ebert highlighted how human rights frameworks should serve as the foundation for ethical AI governance, advocating for a forward-looking approach that asks what kind of societies we want AI to help build. Jan Gerlach emphasized civil society’s dual role as both participants in governance processes and builders of digital public goods like Wikipedia, noting that AI systems are often trained on data from these community-curated sources.
The discussion also addressed persistent digital divides, with Dan York pointing out that 2.6 billion people remain offline, potentially deepening inequalities as AI tools become more prevalent. Panelists called for open standards and protocols in AI development, similar to those that enabled the Internet’s success, while supporting innovation without permission. The session concluded with recognition that balancing AI innovation with societal protection remains a critical challenge requiring continued multi-stakeholder collaboration.
Keypoints
## Major Discussion Points:
– **Multi-stakeholder governance model preservation**: Strong emphasis on defending the distributed, multi-stakeholder Internet governance model against attempts to impose centralized state control, particularly in the context of WSIS Plus 20 review and AI governance frameworks.
– **AI ethics and human rights integration**: Discussion of how to embed human rights frameworks, transparency, and accountability into AI governance, with focus on ensuring AI serves society rather than deepening inequalities or undermining democratic participation.
– **Civil society participation and shrinking civic space**: Concerns about maintaining meaningful civil society engagement in Internet governance processes, including challenges with funding, access to forums, and threats to multi-stakeholder participation in various UN processes.
– **Digital divide and connectivity gaps**: Recognition that 2.6 billion people remain unconnected to the Internet, and that AI development may be widening rather than closing digital divides, excluding voices from global South and marginalized communities.
– **Information integrity and trustworthy ecosystems**: Focus on combating disinformation while protecting freedom of expression, supporting independent journalism, digital public goods like Wikipedia, and ensuring diverse voices are represented in information systems.
## Overall Purpose:
The discussion aimed to explore how the WSIS Plus 20 review process can strengthen inclusive, rights-based digital governance, particularly regarding AI ethics and information integrity. The session sought to develop concrete policy recommendations for maintaining multi-stakeholder engagement while addressing emerging challenges from AI and threats to Internet freedom.
## Overall Tone:
The discussion maintained a collaborative yet urgent tone throughout. Participants expressed shared concerns about threats to the multi-stakeholder model and human rights online, while remaining constructively focused on solutions. There was an underlying tension between optimism about technology’s potential benefits and anxiety about current challenges to Internet governance and civil society participation. The tone became slightly more pressing toward the end when discussing immediate threats like the Open-Ended Working Group negotiations and calls to pause AI regulation for competitive reasons.
Speakers
– **Ernst Noorman** – Ambassador for Cyber Affairs of the Netherlands, Session Moderator
– **Rasmus Lumi** – Director General, International Organization and Human Rights at the Ministry of Foreign Affairs of Estonia, Chair of the Freedom Online Coalition
– **Gitanjali Sah** – Strategy and Policy Coordinator at the ITU, responsible for the World Summit on Information Society process
– **Thibaut Kleiner** – Director for Policy, Strategy, and Outreach at DG Connect of the European Commission, former Head of the Unit of Network Technologies
– **Isabel Ebert** – Senior Advisor of Business and Human Rights and Tech at the Office of the High Commissioner of Human Rights of the UN, Advisor at the BTEC Project on Business and Human Rights in the Technology Sector, member of the OECD AI Group of Experts
– **Jan Gerlach** – Public Policy Director of Wikipedia Foundation, leads global advocacy efforts within Wikimedia Foundation
– **Dan York** – Chief of Staff of the Office of the CEO at the Internet Society, background in DNS, real-time communication, and IETF involvement
– **Participant** – Riyad Abathia, former NGO’s Coordination Office in the United Nations, international civil society activist
Additional speakers:
None identified beyond the speakers names list.
Full session report
# Strengthening Multi-Stakeholder Digital Governance: WSIS Plus 20 Discussion Report
## Introduction
This discussion, moderated by Ernst Noorman, Ambassador for Cyber Affairs of the Netherlands, brought together representatives from international organizations, governments, civil society, and the technical community as part of the World Summit on Information Society (WSIS) Plus 20 review process. The session, organized by the Freedom Online Coalition, focused on artificial intelligence ethics and information integrity within the context of multi-stakeholder Internet governance.
The panel included Rasmus Lumi (Director General, Estonia Ministry of Foreign Affairs and FOC Chair), Gitanjali Sah (ITU Strategy and Policy Coordinator), Isabel Ebert (UN Office of the High Commissioner for Human Rights), Thibaut Kleiner (European Commission DG Connect), Dan York (Internet Society), and Jan Gerlag (Wikimedia Foundation).
## Opening Framework: Freedom Online Coalition Priorities
Rasmus Lumi established the session’s context by highlighting threats to foundational Internet governance principles. He emphasized that “we cannot overcome the challenges without the meaningful engagement of all shareholders” and warned that “efforts to impose centralised control threaten to undermine the Internet’s fundamental openness, risking fragmentation and compromising the very attributes that have made the Internet a catalyst for progress and innovation.”
Lumi positioned 2025 as a critical year when “long-standing values” face challenges, noting that some countries are attempting to “veto multistakeholder civil society organizations out of the room.” The Freedom Online Coalition’s response involves working through multi-stakeholder formats to defend established governance principles.
## Multi-Stakeholder Governance Approaches
### Institutional Perspectives
Gitanjali Sah emphasized that “WSIS Plus 20 provides opportunity for multi-stakeholder dialogue to include all voices in the UN General Assembly review process.” She highlighted the ITU’s commitment to inclusive participation, noting the WSIS Forum’s “literally five months” open consultative process and efforts to accommodate remote participation and different time zones.
Thibaut Kleiner advocated for strengthening existing mechanisms, suggesting that the “Internet Governance Forum should become permanent UN institution with own budget and director for ongoing discussions about emerging technologies.” He emphasized that approaches should be “bottom-up, owned by local constituencies rather than imposed.”
### Technical Community Perspective
Dan York brought technical expertise to the governance discussion, noting that “technical communities bring essential expertise whilst civil society provides knowledge about impacts on vulnerable populations.” He emphasized the Internet’s success through “open standards, open protocols, and innovation without permission principle” while expressing concern about maintaining this balance as governance evolves.
## Human Rights Framework for Digital Governance
Isabel Ebert positioned human rights as the foundation for digital governance, arguing that “human rights framework should serve as common minimum denominator for ethical approach to technology.” She reframed the AI governance debate by suggesting we ask “what kind of societies do we want AI to help us build and which accountability structures for different actors and their distinct role can incentivise this.”
Ebert called for “transparent rules matching pace of AI development with benefits shared across nations and risks thoughtfully managed,” emphasizing transparency, accountability, and inclusion as core principles.
## Artificial Intelligence Governance Challenges
### Cross-Sectoral Coordination
Gitanjali Sah emphasized that “AI governance must be cross-sectoral, looking across health, agriculture, education with overarching ethical framework.” This comprehensive approach recognizes AI’s impact across all sectors of society rather than treating it as a standalone technical issue.
### Technical Standards and Openness
Dan York advocated for “open standards and protocols for AI transparency, explainability, and accountability,” drawing parallels with Internet development. He expressed concern about “proprietary, closed AI systems creating vendor lock-in and concentrated power,” emphasizing the need to maintain open, collaborative approaches in AI development.
## Civil Society as Digital Infrastructure Builders
Jan Gerlag provided a significant reframing of civil society’s role, emphasizing that “civil society, the people who use the internet, also build large parts of the internet… They build the digital public goods that the Global Digital Compact aims to support.” Using Wikipedia as an example, he illustrated how civil society creates and maintains critical Internet infrastructure through “massive self-governed collaboration systems.”
Gerlag noted that “good AI governance requires supporting communities who curate and verify information that feeds AI systems,” highlighting civil society’s role in creating trustworthy information ecosystems. However, he warned that “civil society input is critical for internet governance success, but their access to these processes is under threat.”
## Digital Divides and Connectivity
Dan York introduced sobering statistics, noting that “one-third of the world (2.6 billion people) still lacks internet access, and AI development risks deepening digital divide.” He explained how technological advancement can worsen inequalities: “Those of us who have access to the AI tools and systems that we’re all using, we are able to be more productive… And we’re leaving the folks who are offline further behind.”
York also highlighted that “those without connectivity cannot contribute knowledge to information pools used for training AI models,” showing how digital exclusion affects both access to AI benefits and representation in AI systems.
## Information Integrity and Community Approaches
Jan Gerlag presented Wikipedia as a model for community-driven information integrity, noting that Wikipedia and similar projects “provide vital information access and represent massive self-governed collaboration systems.” The Wikimedia Foundation’s approach emphasizes supporting “individual agency through literacy, privacy, safety and transparency” rather than relying solely on top-down content moderation.
The discussion highlighted the need to “support civil society organisations through smart policies and funding to sustain trustworthy information ecosystems,” recognizing that community-driven approaches require institutional support to remain sustainable.
## International Cooperation and Regulatory Balance
Ernst Noorman addressed tensions between innovation and protection, noting that “right now, if you look at the AI discussion, it’s more and more about competition. Who will be the winner?” He argued that “AI is there to serve society and humanity” and criticized calls to “pause the EU AI Act because of competition reasons,” stating that “competition concerns should not override regulation designed to protect society and create level playing fields.”
## Audience Engagement and Practical Concerns
The session included audience participation, with questions about regional coordination and the role of national chapters in Internet governance. Speakers emphasized leveraging existing mechanisms, including what Dan York noted as “180 different national or regional Internet Governance Forums” worldwide, rather than creating entirely new structures.
## Key Challenges and Ongoing Issues
The discussion identified several persistent challenges:
– Ensuring meaningful multi-stakeholder participation while some actors attempt to exclude civil society
– Balancing AI innovation with societal protection and rights-based approaches
– Addressing the digital divide while preventing AI from deepening existing inequalities
– Maintaining open, collaborative approaches in AI development similar to Internet governance
– Supporting civil society organizations’ capacity to participate in governance processes
## Conclusion
The session demonstrated broad agreement on the importance of multi-stakeholder governance and human rights-centered approaches to digital governance, while revealing different emphases on implementation strategies. The discussion highlighted the need to defend established Internet governance principles while adapting to emerging challenges from AI development and persistent digital divides.
The session concluded with time constraints as “the president of Estonia is about to make his remarks,” reflecting the broader context of high-level diplomatic engagement around these issues. The emphasis throughout was on maintaining inclusive, participatory approaches to governance while ensuring that technological development serves societal needs rather than merely competitive interests.
Session transcript
Ernst Noorman: Good morning everyone. Very much welcome to this session. First of all, my name is Ernst Noorman. I’m the Ambassador for Cyber Affairs of the Netherlands. By the way, I also want to welcome the online participants to this session. Before I introduce the panelists, I will introduce the subject of this morning. As we approach the 20th year review of the World Summit on Information Society, or the WSIS Plus 20 as we all know it, it’s a timely moment to reflect on how we can strengthen inclusive rights-based digital governance. This session focuses on WSIS Action Line C10 and C11 on ethical dimensions of the information society and international cooperation. Our goal is to explore how multi-stakeholder engagement, including civil society, the private sector, academia, and end-users can help shape digital spaces that uphold human rights and support sustainable development. A key part of this conversation will be the role of artificial intelligence, especially generative and decision-making systems, in shaping the integrity of online information, trust, and democratic participation. We’ll look at how governance frameworks can promote transparency, accountability, and equity while protecting freedom of expression, privacy, and non-discrimination. We’ll also consider whether current international and human rights frameworks are equipped to respond to the rapid evolution of AI and how we can work together to prevent these technologies from deepening existing inequalities. Finally, we’ll highlight, at least I hope, practical and collaborative approaches to bridging digital divides and building trustworthy information ecosystems that advance the sustainable development goals. I look forward to an engaging discussion with concrete strategies and policy ideas that can help to shape a more inclusive and ethical digital future. Now, for that, we have five excellent speakers, which I will introduce right now. First of all, we have Gitanjali Sa to my right, and she’s the Strategy and Policy Coordinator at the ITU and is responsible for the World Summit on Information Society process. Then we have Thibaut Kleiner, to the left of me, and recently appointed as Director for Policy, Strategy, and Outreach at DG Connect of the European Commission, and also experienced as before as the Head of the Unit of Network Technologies, and this unit was in charge, or is in charge, of research and innovation in the area of wireless optical networks, network architecture, Internet of Things, SATCOM, and the 5G public-private partnerships. Then we have Isabelle Ebert, to the right of me, Senior Advisor of Business and Human Rights and Tech at the Office of the High Commissioner of Human Rights of the UN, and Isabelle is an Advisor at the BTEC Project on Business and Human Rights in the Technology Sector and a member of the OECD AI Group of Experts. Then we have Jan Gerlag, to the left of me. Jan is the Public Policy Director of Wikipedia Foundation, and at Wikimedia Foundation he leads efforts within the global advocacy teams to educate lawmakers and governments worldwide on Internet policies that promote and protect Wikipedia and open knowledge participation. And finally, as a participant in the panel, is Dan York, and Dan serves as the Chief of Staff of the Office of the CEO at the Internet Society, advising the President and CEO, coordinating organizational priorities and managing key relationships, and his recent work has focused on Internet shutdowns, resilience, and projects such as sustainable technical communities, Leo satellites, and open standards everywhere. And with a background in DNS, real-time communication, and long-standing involvement in the IETF, Dan has been working with online technologies since the mid-80s, so a long experience. But first, we start off with my dear friend Rasmus Lumi, Director General, International Organization and Human Rights at the Ministry of Foreign Affairs of Estonia, and right now the Chair of the Freedom Online Coalition, to share his thoughts on what the role of the FOC and similar initiatives could play in shaping AI and information integrity standards in the WSIS-20 process and beyond. Please welcome to the floor. Thank you.
Rasmus Lumi: Thank you very much, and I’m very glad and honored to be here, to be able to deliver the opening remarks to this very distinguished panel. So, first of all, I’d like to say that it’s kind of obvious, maybe, but I think it is also needed to be repeated that 2025 seems to be the year where our long-standing values and principles are being challenged more than ever, and international organizations, especially the United Nations, are in notable difficulties. While, as always, this presents opportunities, I’m afraid we will much more likely be struggling with the challenges. With this in mind, we will have to renegotiate the future of the Internet management. We all know that the multi-stakeholder format is of key importance here. We cannot overcome the challenges without the meaningful engagement of all shareholders. We need coordinated response. This is where the Freedom Online Coalition comes to play. The Freedom Online Coalition’s core mission to promote human rights and fundamental freedoms online remains essential, and this roundtable reflects FOC’s commitment to ensuring that digital transformation is rights-based, and the FOC is a good example of a multi-stakeholder format. As we approach the WSIS Plus 20 review process, we, the like-minded, both through the FOC and in other ways, must unite our efforts to resist any attempts to overturn the existing distributed multi-stakeholder Internet governance model and attempts to expand state control over the Internet. We must adopt a strong common approach to ensure the protection of the Internet’s decentralized model. Efforts to impose centralized control threaten to undermine the Internet’s fundamental openness, risking fragmentation and compromising the very attributes that have made the Internet a catalyst for progress and innovation. Multi-stakeholder approach enshrined in the Tunis agenda and reaffirmed in the GDC is based on the premise that effective Internet governance must be inclusive, participatory, and consensus-driven, involving a broad array of actors from the public sector, private sector, civil society, technical community, academia, regional and international organizations. Multilateralism alone is not sufficient to solve the global digital challenges. Given what I said before about attempts to overthrow the current Internet governance model, it is a risk. Multi-stakeholder models ensure that all relevant actors, including the technical community and so on, are part of the open conversation. Deeper collaboration between the stakeholders is more important than ever to address cross-border challenges. We have to integrate multi-stakeholder involvement into multilateral forums. The continued general availability and integrity of the Internet as a global and interoperable network of networks is fully dependent on the continued functioning of the multi-stakeholder model. This session is an opportunity to develop concrete recommendations that can complement both the WSIS plus 20 process and the FOC’s ongoing efforts to uphold Internet freedom worldwide. Finally, as a token of appreciation to the multi-stakeholders, I would like to thank the Freedom Online Coalition’s advisory network for their proactive advice on WSIS plus 20, as well as on the Elements paper. We will work together with FOC member states to take this feedback into account in our national positions. Thank you very much.
Ernst Noorman: Thank you very much, Rasmus, for your opening words. And also for your leadership this year in the FOC, the Freedom Online Coalition. Now, with the question to the panellists, let me, allow me to start with you, Gitanjali Sah. How do you see that WSIS plus 20 can advance ethical and rights-based digital governance, particularly in the context of AI and information integrity?
Gitanjali Sah: Thank you so much, moderator. Good morning, everyone. It’s nice to see a full room because it’s such an important topic. You know, going forward, we need opportunities of multi-stakeholder dialogue like we have, we got at the IGF, we’re getting here at the WSIS forum, so that all the voices are included and put forth to the UNGA overall review. This is where the decisions will be made in December. They will come out with an outcome document where, which should reflect the urgencies, especially the urgencies of the ethical dimension that we face right now. So within the WSIS process, we do have an action line on cyber security as well, along with ethics and access. It’s all cross-sectoral. So what we heard out here is, even when we are talking about regulation, we have to look across all sectors. We had a regulators round table, which, which really concluded with this aspect that, one, we have to look across We need more best practices that we can share across all the countries, across all the stakeholders. Secondly, we really need to have more cross-sectoral work. What we are talking about in health is also equally important in agriculture, is equally important in education. So when we are talking about AI ethics, we really need to ensure that we are looking at it from an overarching framework and it is cross-sectoral. The third thing that we have been pressing upon is that other than the awareness that we are creating with organisations like the, really not an organisation but a group, Freedom Online Coalition, for us like this, awareness building is extremely important so that not only the regular stakeholder communities but also communities of educators, communities of really engineers, the private sector that is designing all of this has this moral responsibility and is included in these kind of discussions. So yes, awareness, ensuring all communities are involved as the UN system, as ITU, we are committed to providing a platform, an equal and just platform where all stakeholders can have a voice and that the voice is included in the UN processes. So really going forward, the ethical dimension remains a crucial element, rights of people online. We were just discussing the rights of children also in the previous session that, you know, there is a lot happening on the internet, the dark net, how do the children know about their rights, are the schools educating them, do we have the right governance structures, do we have guidelines for parents, for educators, so a lot of work has to be done, but the fact that we are discussing it and we are making sure that it’s inputted into the UN process, the overall UN process through WSIS is very important. Thank you very much.
Ernst Noorman: Thank you very much, Gitanjali, for your comments. Isabel, now what do you see as the most effective way to promote transparency, accountability and inclusion in AI governance through multi-stakeholder cooperation?
Isabel Ebert: Thanks very much. Thanks a lot also to the Freedom Online Coalition to continue to convene this very important dialogue. It’s an honour to be here. Yeah, I think what we see, I mean also the conference taking place very closely to here, the core goal of this discussion is that we want to realise that technology can really power benefits, but in order to do so we need to make sure that people are not left behind and we are not undermining the purpose of the technology that it sets out to achieve by ignoring certain risks. And here it’s really important also to bring in the ethical dimension and international cooperation that is key to achieve that we are able to realise the benefits by managing the risks. Here there’s always this debate around ethics, human rights, how does it act together. Human rights framework as such is the framework that the member states of the UN have committed to, that the Global Digital Compact has endorsed and it should really serve as the common minimum denominator to conceptualise whether it’s an ethical approach to technology. The other aspect I wanted to highlight is that a responsible technology future is not automatic by just applying technologies to all spheres of society. We need to first understand what type of responsible technology future we want and then see how technology can support this. And these choices we are making in the WSIS process are really important to ensure that the parameters for AI governance we are setting are responsible and rights respecting. So with regard to transparency, accountability and inclusion I would like to lead with three reflections. Firstly, creating transparent rules of the road are important by introducing policies and oversight mechanisms that can match the pace and scope of AI development, new technologies development and ensuring that the benefits are communicated and shared across nations and that risk to people are thoughtfully managed and anticipated. Second, we need to shift the terms of the debate to a forward-looking and solution-oriented accountability conception. Instead of asking how do we adapt AI, we should ask what kind of societies do we want AI to help us build and which accountability structures for different actors and their distinct role can incentivise this. Not only through regulation but also through incentive-based stimulus packages. And thirdly, making the rules of the game inclusive. So with regard to exclusivity, it’s important to expand who gets to participate in the decisions around AI governance beyond state, equipping multilateralism also with a dialogue with affected communities that are often not sufficiently reflected in these processes and making sure that also the design processes of new technologies are developed in engagement with communities in order to make better products, safer products, which also again then bringing me to the most important stakeholder group, at least for BTEC, ensuring responsible business conduct, ensuring that the companies that are at the forefront of developing new technologies build human rights into their products and services. And in that regard, human rights provide a guidance how AI can be governed. The UN Guiding Principles on Business and Human Rights define respective roles and responsibilities of states and companies towards working human rights and they are also very important to, since they hadn’t been in place when WSIS was conceived initially, to now take into account for the very important process this year. Which leads me to conclude that the multi-stakeholder model is really essential to achieve transparency, accountability and inclusion in AI governance and the human rights frameworks, both the International Human Rights Framework as well as the UN Guiding Principles on Business and Human Rights can help us to understand what is at stake in the AI governance debate. The BTEC project has published a taxonomy of how AI relates to human rights. We are putting out a lot of guidance in order to, what I said earlier, promote the solutions-oriented, forward-looking accountability approach and this applies to a range of rights, non-discrimination, privacy, access to information, freedom of expression and we are convinced that innovation on human rights values can deliver a lot of benefit to the people.
Ernst Noorman: Thank you very much. Jan, I think I have an excellent question for you as a follow-up also to the remarks of Isabel and that’s what you see as a role that civil society should play in ensuring that the WSIS plus 20 process leads to a rights-based AI governance and the development of trustworthy inclusive information ecosystem, especially in the face of disinformation and shrinking civic space.
Jan Gerlach: Thank you, Ernst, for the question. It’s quite complex, it took me a bit to unpack it first but I think from a Wikimedia perspective I’d say there are two main things to note about civil society participation in this process. First, and this one may sound obvious, but civil society input makes internet governance and regulation better. In fact, it’s critical for our shared success. However, we’re fighting a bit and maybe that’s an understatement right now for the future of civil society access to internet governance processes. Civil society’s ability to participate in these conversations, just like this one, is directly affected by the outcomes of the WSIS plus 20 review and with it its ability to inform future processes like this about trustworthy and inclusive information ecosystems. Now the FOC’s blueprint for information integrity, which Wikimedia, by the way, contributed to, supports individual agency through literacy, privacy, safety and transparency. The blueprint also promotes trust through, again, transparency and accountability around platforms at work, around their products, around algorithms that are used through support for reliable sources of information, including independent journalism, digital public goods, through privacy and safety, especially that safety of marginalized and vulnerable groups. And finally, the blueprint seeks to promote inclusion through linguistic and cultural diversity, through meaningful connectivity, the promotion of diverse and global voices and protection against discrimination and harassment. In all of this, people are front and center, individuals because the internet really must serve them. Now, not all internet users can fit in this room, let alone in smaller ones, and not every perspective will be represented by the members of civil society here. So those people from civil society organizations who are here and who can travel here really need to work hard to ensure that the values and measures of the blueprint for information integrity are turned into reality and informed policy at the international, at the national and local levels. That requires a lot of coordination among civil society groups and organizations, and also a lot of engagement with their own stakeholders to ensure the needs of people around the world are properly understood and fed into these processes. The conversations we’re having here and in other places around the world about the future of WSIS, about the future of IGF and multi-stakeholderism in general, must have civil society participation in order to include voices that otherwise wouldn’t be heard. These voices make internet governance better, especially during times of shrinking civic spaces, and they help ensure that AI governance, too, truly serves people and their rights when we are all at the risk of really drowning in disinformation and synthetic information that isn’t verifiable. Now to my second point, the thing I want to talk about is that civil society, the people who use the internet, also build large parts of the internet. I think that’s underappreciated. They build the digital public goods that the Global Digital Compact aims to support. They are the independent journalists that the blueprint for information integrity wants to support. They are the people building all the small open knowledge projects that underpin the free and open Internet that we all like to talk about. Take Wikipedia as an example, probably the most prominent example of such projects. This online encyclopedia is built by thousands of volunteers from all around the world, from all walks of life, who contribute their time to add content, engage in conversations about the policies that govern Wikipedia, etc. Wikipedia provides vital access to information for populations around the world, and it’s a massive self-governed system of collaboration from people from everywhere. To ensure such projects can continue to thrive, these people, through civil society organizations, need to have a voice at the table of Internet governance processes. One other point I want to make is this. In the halls of AI4Good, there’s a lot of impressive technology on display, mostly built by the private sector, but the science behind it often stems from work in academia. And there are some academic booths, too, and it’s a good reminder that academia really needs to be part of these conversations as well. Lastly, a lot of AI systems, large language models, etc., are trained on data that comes from projects like Wikipedia that verify knowledge and update information regularly. This open ecosystem of trustworthy information needs to be sustained. Good AI governance, among other things, means to support the communities who curate and verify the information that feeds AI. If we want to support a trustworthy ecosystem of information in the age of AI, governments, including FOC member states, must make sure to support this part of civil society as well, through smart policies and through funding. Thank you.
Ernst Noorman: Thank you very much, Jan. You referred also to different processes going on, and where the multistakeholder involvement is so crucial. It makes me also think of right now the Open-Ended Work Group, which takes place right now in New York. This week is a crucial week in New York on the follow-up of the Open-Ended Work Group on responsible state behavior in ICTs. Multistakeholder involvement there is on top of the agenda for us, for many of us like-minded countries, and it has an especially difficult position there, because many countries want to veto multistakeholder civil society organizations out of the room, and unfortunately have been quite successful at that as well. So you can also ensure that, especially from the few FOC countries, they will be fighting for the future and the follow-up of the Open-Ended Work Group to ensure multistakeholder involvement. But let me continue with Thibaut. How do you see that international and regional and multistakeholder partnerships can help bridge digital divides and support trustworthy, inclusive information ecosystems in the context of WSIS and beyond?
Thibaut Kleiner: So first of all, I’d like to congratulate the organizers for having this topic here so prominently also in the WSIS conference. I think that indeed these days, as you pointed out, the risks linked with technology towards human rights are just growing. The ability of AI, of surveillance technologies to infringe on human rights have just increased, and therefore it’s very important that we do not lower the attention. On the contrary, we should make more attention to this. And in the context of WSIS, I think the European Union has been, and it was very clear also when we were negotiating collectively on the Global Digital Compact. I mean, we’ve really insisted that these human rights dimensions cannot be neglected. They have to be forefront. They have to be really at the heart of what we are talking about. And in a way, this is something we have reflected in the recent past with our declaration on digital rights and principles. That’s really something that we’ve tried to encompass, these various elements along six pillars that we believe reflect very much the types of challenges we have also for WSIS. It’s about putting people and their rights at the center of the digital transformation. It’s about supporting solidarity and inclusion, ensuring freedom of choice online, fostering participation in the digital public space, increasing safety, security, and empowerment of individuals, in particular children, and promoting the sustainability of the digital future. And the interesting thing is that this declaration actually has been also underpinning the regulatory efforts that we have conducted in the EU, developing indeed hard elements and obligations towards private and public actors. And this is, I think, what we can now, through WSIS, organize in terms of discussions between multi-stakeholders. And very much in our view, we want to make sure that the Internet Governance Forum towards WSIS becomes a permanent institution with its own budget from the UN with a director, and also that it can become really the place where we have also repeatedly discussions about how the evolving, the emerging technologies can be looked at, and so that we can make sure that we don’t overlook the importance of protecting human rights. And within WSIS, we think that indeed regions, but also countries and even locally, we can engage in these conversations. And this is what we have tried to do also through various projects the EU is supporting, trying indeed to engage with countries in Africa and Latin America, so that we not only explain the challenges of the digital technologies and the risks that I implied, but also so that we support public debate in these regions. Because at the end of the day, it is not something you can impose from the outside. It is very much our belief that human rights is not something that is just coming from certain countries or regions globally. It’s something that is universal, and it’s something that we very much believe can be bottom-up, owned by the local constituency. So our approach is very much not to impose this view, but actually to try and promote discussion and dialogue with our various partners internationally. And as I said, we try to highlight the way we see it, but we very much believe that it is about communities, it is about companies, the public sector, but also the youth embracing these elements. And I think that’s what WSIS can achieve, really creating a space for dialogue and making sure that we not only put human rights as nice to have, but actually centerpiece for everything we try to build.
Ernst Noorman: Thank you very much, Thibaut. Then I move to my right, the far furthest right, Dan. With all your experience, how can technical infrastructure, internet standards and governance protocols be strengthened to support trustworthy information ecosystems and ethical AI deployment? And what role should multistakeholder cooperation play in this effort?
Dan York: Thank you for that question, and thank you for the Freedom Online Coalition for hosting this session today. And as a 20-year editor of Wikipedia, I want to say thank you to the Wikimedia Foundation for all that they do around here. So the Internet Society was founded in 1992 by a collection of civil society, academics, technical universities, internet companies, to really build an organization and to think about this vision that the internet is for everyone. It was also the home of the Internet Engineering Task Force, or IETF, which has been the standards organization, has brought us stuff like TCPIP, HTTP, the protocols that allow us to work. And I think if we look at the last 20 years of internet governance, of internet operations and pieces, and the lessons that can be applied forward toward what we’re looking at now with AI and the ethics of AI, et cetera, one of those key elements is the importance of open standards and open protocols and the open development of those standards and pieces that are there. That’s really what got us to where we are. And I think a concern we see from the technical side is that we’re seeing a lot of interest in more proprietary, closed AI systems, et cetera, which create the same kind of issues that we see in some parts of the internet today. Vendor lock-in, closed proprietary systems, concentrated power, those kind of things. Just the internet protocols like TCPIP and HTTP, just as they enable global interoperability, we need to also think about what standards can there be for AI transparency, AI explainability, AI accountability. There are some standards starting to be developed by different groups within some parts of that, but it has to happen at all layers of the AI stack, as we might refer to it as, in some kind of form. And those standards need to be developed in a multi-stakeholder way. You know, the technical communities bring essential knowledge and expertise about system design. Civil society brings important, you know, how those AI systems impact vulnerable populations in ways that we may not necessarily grasp. Governments bring information about policies and how these can be extended in ways. End users, you know, provide crucial feedback about how they can change and shape. These are the elements that need to be all part of it. AI systems and policies need to be developed by, you know, the rural farmer that we are. and the students who might be using AI-affected agriculture, and also the students who might be using AI-moderated education. It’s something that’s there. And another key point for us is that we have to think about the fact that a third of the world is still not connected. There’s 2.6 billion people who do not have access to the internet. And those who do don’t necessarily have affordable, reliable, or resilient connectivity. We’re, in fact, with some of what we’re doing, we are deepening the digital divide. Because those of us who have access to the AI tools and systems that we’re all using, we are able to be more productive or use things in different ways. And we’re leaving the folks who are offline further behind. And also, we’re not gaining access to the knowledge and information that they may have. They are not contributing into the pools of information, as was mentioned earlier, that are being used to train these models and to work with things. So there’s a need to bring that in as well. So it’s this combination of open standards, open protocols, ways to involve everyone and connect those who are there, and to really bring all of the folks in to be part of this. And I think the last piece I would just mention is that one of the principles of the internet that has made it work so well to where we are today is this idea that you can have innovation without permission. The ability to go and create new ideas, bring things out, publish new reports, publish new websites, open up new tools, without having to go and ask somebody permission or pay somebody to put this online. Some of us who may have been around before the internet remember a time when you couldn’t put anything online unless you paid somebody to do so. It was a different world. We need to figure out how to balance so that we are protecting the harms and things, but also ensuring that that level of innovation continues and to work with that. So a bit of that from the technical community side, and we’re looking forward to working with all of you in trying to help as we continue to move out into this new world.
Ernst Noorman: Thank you very much. We do have still some time for questions. I do not see actually microphones. I see at least a hand, so that’s very good. But now the question is if you have a microphone to raise your question. Otherwise you… Someone is running. Yes, great. Please, introduce yourself and…
Participant: Thank you very much. My name is Riyad Abathia. I’m a former NGO’s Coordination Office in the United Nations, international civil society activist in the UN since more than 20 years ago. I’m contributing, we are contributing in WSIS process since the beginning, and civil society effort is recognizable in the WSIS process since the beginning. Ten years earlier, we are celebrating this year 20 years, the best age of life. Ten years earlier, ten years later, the state member adopted precious document and it’s working since then like that. Some other foundation NGOs are actively engaged in the WSIS process. We have a largest network for cities, largest workers for spectrum. We talk spectrum and quick term and other NGOs also. What I want to say that… My questions, yes. I appreciate your talk. But anyway, those institutions who are not international organizations, who are not non-governmental organization, who are not foundation, and we have at least one representative in the panel, ISOC, Internet Society. You are like IEEE, you are like IGF, you are like top level demand, you are five or seven institutions. Your contributions in the Internet community is very highly determinated. But I don’t know because ITU, what they’re doing with state members and the complement effort of civil society. But until now, as Secretary General said the first day, that two billions didn’t be connected until now. But yes, ISOC Congress hosted in Geneva 2012, it was largest 3,000 delegates. But how about the enhancing of regional coordination offices of ISOC, for example? How about to support the national office chapters of ISOC? This is what is needed also and state members are expecting and also ITU will be complemented. Thank you very much.
Ernst Noorman: Hopefully my question has been clear. That’s a question to Dan. Yes. Okay.
Dan York: So, I think you’re asking how can we, as the Internet Society and other groups like that, help these discussions at a national level or local level, regional level? Yeah, I think, so thank you for asking that question. The Internet Society does have 120 chapters around the world in various different areas. Some of those are national, some regional in different areas. And those chapters are very engaged in these kind of conversations, perhaps not in some of this level of AI ethics and pieces around that because this is a newer topic. In many areas, we’re mostly focused around the connectivity and in trying to ensure that people have affordable, resilient, reliable, you know, connectivity and pieces. But this is part of the bigger picture. Some of our chapters are very involved with AI topics and other elements around that. So, it varies widely because they are individual organizations. I think the bigger picture, though, is that how do you engage people from out in various different regions and areas who cannot necessarily come into these forums? The comment earlier about the civil society aspect of this, it’s hard for civil society organizations and others to participate in venues such as this. One of our concerns, certainly, is that we would like to see, we don’t want to see a proliferation of more events or more mechanisms. Because each one of those is more cost, is more elements that it’s harder for organizations to be able to participate in. So, that is one concern. But we certainly do see that organizations at a national level, at a regional level, should be able to have some way to participate in these kind of venues, which is why we’re a big fan of the Internet Governance Forum and the way that it brings people together. There are now about 180 different national or regional Internet Governance Forums which are bringing people together all around the world. And those are other elements and ways that people are having a voice into the ongoing conversations that we’re having. So, thank you again for the question.
Ernst Noorman: Thank you for the response. You want to react?
Gitanjali Sah: Yes, thank you. Riyad, we have ensured as ITU that the WSIS remains multi-stakeholder. For instance, even at the WSIS Forum, you have remote participation in every room for the civil society organizations who couldn’t be here. We are also very responsive about the requirements of the regional time zones. So, we try to accommodate that as well. And as you all know, the agenda and the program of the WSIS Forum is built through an open consultative process. It’s really like literally five months where you can input through an official form and let us know what you want to see at the WSIS Forum. And civil society is a very active partner for that. We have several physical meetings as well. And we really want you to contribute and to help us to ensure that we keep this dialogue going beyond 2025 as well. I just also wanted to add, moderator, that when the WSIS outcome documents were drafted, they were really a universal declaration of human rights is right on the second page of the WSIS outcome document, right? So, we must also recall that the documents were drafted in a very, very inclusive and sound manner. And we must continue that spirit of inclusion, of making sure that the framework of the WSIS Action Lines continue to remain relevant as they have evolved with the technological changes. We would have to finish, moderator, because the president of Estonia is about to make his remarks here in this room.
Ernst Noorman: Okay. Then… So, great to have a timekeeper. We have indeed two minutes, but I’m afraid it’s too short for another question and then responses. What are my concluding remarks is that this discussion is still so especially relevant on how to involve multistakeholder. We also have to look, I think, at what do we mean all with multistakeholders? How do we balance the different voices in the multistakeholder model? And also, right now, if you look at the AI discussion, it’s more and more about competition. Who will be the winner? And well, as Isabel, amongst others, said, you know, AI is there to serve society and humanity. And how do we ensure that? And just lately, just these days, you can hear even in Europe, calls to pause the EU AI Act because of competition reasons. I think that’s not the good way to go. Regulation is also there to protect our society. And so far, we have seen often that regulation can even improve innovation by creating a level playing field. So, in that sense also, we have to fight and see how we can continue to involve the multistakeholders in all levels of the discussion. So, let’s go for that. And again, also, Rasmus, I want to thank you and lots of success in your continued work for the next six months as the chair of the FOC. And thank all panelists for your contributions. Thank you very much.
Rasmus Lumi
Speech speed
141 words per minute
Speech length
516 words
Speech time
218 seconds
Need to resist attempts to overturn distributed multi-stakeholder Internet governance model and prevent expansion of state control
Explanation
Lumi argues that like-minded countries must unite through the Freedom Online Coalition and other means to resist attempts to impose centralized control over the Internet. He warns that efforts to expand state control threaten the Internet’s fundamental openness and risk fragmentation.
Evidence
References the current difficulties at international organizations, especially the United Nations, and mentions the Open-Ended Work Group where many countries want to exclude civil society organizations
Major discussion point
Multi-stakeholder Internet Governance and WSIS Plus 20
Topics
Legal and regulatory | Human rights
Multi-stakeholder approach must be inclusive, participatory, and consensus-driven involving public sector, private sector, civil society, technical community, and academia
Explanation
Lumi emphasizes that the multi-stakeholder approach enshrined in the Tunis agenda and reaffirmed in the Global Digital Compact requires broad participation from all relevant actors. He argues that multilateralism alone is insufficient and that deeper collaboration between stakeholders is essential for addressing cross-border challenges.
Evidence
References the Tunis agenda, Global Digital Compact, and the premise that effective Internet governance must involve a broad array of actors including regional and international organizations
Major discussion point
Multi-stakeholder Internet Governance and WSIS Plus 20
Topics
Legal and regulatory | Human rights
Agreed with
– Gitanjali Sah
– Isabel Ebert
– Jan Gerlach
– Dan York
Agreed on
Multi-stakeholder approach is essential for effective Internet governance and AI governance
2025 presents challenges to long-standing values with international organizations facing difficulties
Explanation
Lumi observes that 2025 appears to be a year where fundamental values and principles are being challenged more than ever before. He notes that international organizations, particularly the United Nations, are experiencing notable difficulties in their operations.
Evidence
General observation about the current state of international relations and organizational challenges
Major discussion point
International Cooperation and Regulatory Approaches
Topics
Legal and regulatory | Human rights
Freedom Online Coalition provides coordinated response through multi-stakeholder format
Explanation
Lumi positions the Freedom Online Coalition as an example of effective multi-stakeholder engagement that can provide coordinated responses to digital governance challenges. He emphasizes the FOC’s core mission to promote human rights and fundamental freedoms online.
Evidence
References the FOC’s advisory network providing proactive advice on WSIS Plus 20 and the Elements paper
Major discussion point
International Cooperation and Regulatory Approaches
Topics
Human rights | Legal and regulatory
Agreed with
– Gitanjali Sah
– Isabel Ebert
– Thibaut Kleiner
Agreed on
Human rights framework should be central to digital governance and AI development
Gitanjali Sah
Speech speed
154 words per minute
Speech length
720 words
Speech time
279 seconds
WSIS Plus 20 provides opportunity for multi-stakeholder dialogue to include all voices in the UN General Assembly review process
Explanation
Sah emphasizes that forums like the Internet Governance Forum and WSIS Forum provide crucial opportunities for multi-stakeholder dialogue where all voices can be included and fed into the UN General Assembly overall review. She stresses that decisions will be made in December with an outcome document that should reflect current urgencies, especially ethical dimensions.
Evidence
References the upcoming December UN General Assembly review and outcome document process
Major discussion point
Multi-stakeholder Internet Governance and WSIS Plus 20
Topics
Legal and regulatory | Human rights
Agreed with
– Rasmus Lumi
– Isabel Ebert
– Jan Gerlach
– Dan York
Agreed on
Multi-stakeholder approach is essential for effective Internet governance and AI governance
AI governance must be cross-sectoral, looking across health, agriculture, education with overarching ethical framework
Explanation
Sah argues that AI ethics discussions must take a cross-sectoral approach, recognizing that issues in health are equally important in agriculture and education. She emphasizes the need for an overarching framework rather than siloed approaches to AI governance.
Evidence
References a regulators roundtable that concluded with the need for cross-sectoral work and mentions WSIS action lines on cybersecurity, ethics, and access
Major discussion point
AI Ethics and Human Rights Framework
Topics
Legal and regulatory | Human rights
Need more cross-sectoral work and best practices sharing across countries and stakeholders
Explanation
Sah advocates for increased sharing of best practices across all countries and stakeholders, emphasizing that solutions and approaches should be shared broadly rather than developed in isolation. This includes ensuring cross-sectoral coordination in addressing digital challenges.
Evidence
References conclusions from a regulators roundtable and the need for overarching frameworks
Major discussion point
Digital Divides and Connectivity Challenges
Topics
Development | Legal and regulatory
Agreed with
– Thibaut Kleiner
– Dan York
– Participant
Agreed on
Digital divides and connectivity challenges must be addressed
Awareness building is crucial for educators, engineers, and private sector designing AI systems
Explanation
Sah stresses that awareness building extends beyond regular stakeholder communities to include educators, engineers, and private sector actors who design AI systems. She emphasizes that these groups have moral responsibility and should be included in ethical discussions about AI development.
Evidence
References discussions about rights of children online, dark net issues, and the need for guidelines for parents and educators
Major discussion point
Civil Society Role and Information Integrity
Topics
Sociocultural | Human rights
Remote participation and regional time zone accommodation help include civil society organizations who cannot attend physically
Explanation
Sah explains that ITU ensures WSIS remains multi-stakeholder by providing remote participation in every room for civil society organizations who cannot attend physically. They also accommodate different regional time zones and use an open consultative process for agenda building.
Evidence
References the five-month open consultative process for WSIS Forum agenda building and physical meetings with civil society
Major discussion point
Civil Society Role and Information Integrity
Topics
Development | Legal and regulatory
Universal Declaration of Human Rights featured prominently in original WSIS outcome documents
Explanation
Sah reminds participants that the Universal Declaration of Human Rights appears on the second page of the WSIS outcome documents, emphasizing that these documents were drafted in an inclusive and sound manner. She argues for continuing this spirit of inclusion while keeping the WSIS Action Lines relevant to technological changes.
Evidence
References the specific placement of the Universal Declaration of Human Rights in the WSIS outcome documents
Major discussion point
International Cooperation and Regulatory Approaches
Topics
Human rights | Legal and regulatory
Agreed with
– Rasmus Lumi
– Isabel Ebert
– Thibaut Kleiner
Agreed on
Human rights framework should be central to digital governance and AI development
Isabel Ebert
Speech speed
164 words per minute
Speech length
700 words
Speech time
256 seconds
Human rights framework should serve as common minimum denominator for ethical approach to technology
Explanation
Ebert argues that the human rights framework, which UN member states have committed to and the Global Digital Compact has endorsed, should serve as the foundational standard for determining ethical approaches to technology. She emphasizes that this framework provides the basis for conceptualizing responsible technology development.
Evidence
References UN member state commitments and Global Digital Compact endorsement of human rights framework
Major discussion point
AI Ethics and Human Rights Framework
Topics
Human rights | Legal and regulatory
Agreed with
– Rasmus Lumi
– Gitanjali Sah
– Thibaut Kleiner
Agreed on
Human rights framework should be central to digital governance and AI development
Need to shift debate to asking what kind of societies we want AI to help build rather than just adapting to AI
Explanation
Ebert advocates for a fundamental shift in how we approach AI governance, moving from reactive adaptation to proactive visioning. She argues we should first determine what kind of responsible technology future we want, then see how technology can support that vision, rather than simply adapting to whatever AI systems are developed.
Evidence
Emphasizes forward-looking and solution-oriented accountability conception
Major discussion point
AI Ethics and Human Rights Framework
Topics
Human rights | Legal and regulatory
Need transparent rules matching pace of AI development with benefits shared across nations and risks thoughtfully managed
Explanation
Ebert calls for creating transparent governance rules that can keep pace with rapid AI development while ensuring benefits are communicated and shared internationally. She emphasizes the importance of thoughtfully managing and anticipating risks to people through appropriate policies and oversight mechanisms.
Major discussion point
Transparency, Accountability and Inclusion in AI Governance
Topics
Legal and regulatory | Human rights
Agreed with
– Jan Gerlach
– Dan York
Agreed on
Need for transparency, accountability and inclusion in AI governance
Making rules inclusive requires expanding participation beyond states to include affected communities in AI governance decisions
Explanation
Ebert argues for expanding participation in AI governance beyond traditional state actors to include affected communities that are often not sufficiently represented in these processes. She emphasizes the importance of engaging communities in the design processes of new technologies to create better and safer products.
Evidence
References the need to equip multilateralism with dialogue with affected communities
Major discussion point
Transparency, Accountability and Inclusion in AI Governance
Topics
Human rights | Legal and regulatory
Agreed with
– Rasmus Lumi
– Gitanjali Sah
– Jan Gerlach
– Dan York
Agreed on
Multi-stakeholder approach is essential for effective Internet governance and AI governance
Disagreed with
– Dan York
Disagreed on
Scope and focus of multi-stakeholder engagement
Companies must build human rights into their products and services following UN Guiding Principles on Business and Human Rights
Explanation
Ebert emphasizes that companies at the forefront of developing new technologies must integrate human rights considerations into their products and services. She references the UN Guiding Principles on Business and Human Rights as defining respective roles and responsibilities of states and companies, noting these principles weren’t in place when WSIS was initially conceived.
Evidence
References UN Guiding Principles on Business and Human Rights and BTEC project’s taxonomy of how AI relates to human rights
Major discussion point
Transparency, Accountability and Inclusion in AI Governance
Topics
Human rights | Legal and regulatory
Jan Gerlach
Speech speed
153 words per minute
Speech length
793 words
Speech time
310 seconds
Civil society input is critical for internet governance success, but their access to these processes is under threat
Explanation
Gerlach argues that civil society participation makes internet governance and regulation better and is critical for shared success. However, he warns that civil society’s ability to participate in these conversations is directly affected by WSIS Plus 20 outcomes and that they are fighting for the future of civil society access to internet governance processes.
Evidence
References the FOC’s blueprint for information integrity that Wikimedia contributed to
Major discussion point
Multi-stakeholder Internet Governance and WSIS Plus 20
Topics
Human rights | Legal and regulatory
Agreed with
– Rasmus Lumi
– Gitanjali Sah
– Isabel Ebert
– Dan York
Agreed on
Multi-stakeholder approach is essential for effective Internet governance and AI governance
Blueprint for information integrity supports individual agency through literacy, privacy, safety and transparency
Explanation
Gerlach outlines how the FOC’s blueprint for information integrity promotes individual agency through multiple mechanisms including literacy, privacy, safety and transparency. The blueprint also promotes trust through transparency and accountability around platforms and their algorithms, and supports reliable information sources including independent journalism and digital public goods.
Evidence
References the blueprint’s support for independent journalism, digital public goods, privacy and safety for marginalized groups, linguistic and cultural diversity, and protection against discrimination
Major discussion point
Transparency, Accountability and Inclusion in AI Governance
Topics
Human rights | Sociocultural
Agreed with
– Isabel Ebert
– Dan York
Agreed on
Need for transparency, accountability and inclusion in AI governance
Civil society builds large parts of internet including digital public goods that Global Digital Compact aims to support
Explanation
Gerlach emphasizes that civil society organizations and internet users don’t just participate in governance discussions but actually build significant portions of the internet infrastructure. He argues they create the digital public goods that the Global Digital Compact aims to support and are the independent journalists that information integrity blueprints want to support.
Evidence
References Wikipedia as an example built by thousands of volunteers worldwide, and mentions small open knowledge projects that underpin the free and open Internet
Major discussion point
Civil Society Role and Information Integrity
Topics
Infrastructure | Development
Wikipedia and similar projects provide vital information access and represent massive self-governed collaboration systems
Explanation
Gerlach uses Wikipedia as a prominent example of civil society-built internet infrastructure, describing it as built by thousands of volunteers from around the world who contribute content and engage in policy discussions. He emphasizes that Wikipedia provides vital access to information globally and represents a massive self-governed system of collaboration.
Evidence
Describes Wikipedia’s volunteer-based model and global reach, noting it’s built by people from all walks of life
Major discussion point
Civil Society Role and Information Integrity
Topics
Sociocultural | Infrastructure
Good AI governance requires supporting communities who curate and verify information that feeds AI systems
Explanation
Gerlach argues that since many AI systems and large language models are trained on data from projects like Wikipedia that verify knowledge and update information regularly, good AI governance must include supporting these communities. He emphasizes that governments, including FOC member states, must support this part of civil society through smart policies and funding.
Evidence
References how AI systems are trained on data from projects like Wikipedia and the need for sustaining open ecosystems of trustworthy information
Major discussion point
Civil Society Role and Information Integrity
Topics
Human rights | Legal and regulatory
Thibaut Kleiner
Speech speed
139 words per minute
Speech length
605 words
Speech time
260 seconds
Internet Governance Forum should become permanent UN institution with own budget and director for ongoing discussions about emerging technologies
Explanation
Kleiner advocates for institutionalizing the Internet Governance Forum as a permanent UN institution with dedicated budget and leadership. He argues this would create a stable platform for repeated discussions about how emerging technologies can be evaluated while ensuring human rights protection remains central.
Major discussion point
Multi-stakeholder Internet Governance and WSIS Plus 20
Topics
Legal and regulatory | Human rights
European Union’s declaration on digital rights and principles puts people and rights at center of digital transformation
Explanation
Kleiner describes the EU’s declaration on digital rights and principles as reflecting six key pillars that address WSIS-type challenges. These include putting people and rights at the center, supporting solidarity and inclusion, ensuring freedom of choice online, fostering participation in digital public space, increasing safety and empowerment, and promoting sustainability.
Evidence
References how this declaration has underpinned EU regulatory efforts, developing hard obligations for private and public actors
Major discussion point
AI Ethics and Human Rights Framework
Topics
Human rights | Legal and regulatory
Agreed with
– Rasmus Lumi
– Gitanjali Sah
– Isabel Ebert
Agreed on
Human rights framework should be central to digital governance and AI development
Disagreed with
– Ernst Noorman
Disagreed on
Approach to AI regulation and competition concerns
Human rights approach is universal and should be bottom-up, owned by local constituencies rather than imposed
Explanation
Kleiner emphasizes that the EU’s approach to promoting human rights in digital governance is not to impose views from outside, but to promote discussion and dialogue with international partners. He argues that human rights are universal but should be embraced by local communities, companies, public sector, and youth rather than imposed externally.
Evidence
References EU projects supporting engagement with countries in Africa and Latin America to promote public debate about digital technology challenges
Major discussion point
Transparency, Accountability and Inclusion in AI Governance
Topics
Human rights | Development
EU supports projects engaging with Africa and Latin America to promote public debate about digital technology challenges
Explanation
Kleiner describes how the EU supports various projects that engage with countries in Africa and Latin America, not just to explain digital technology challenges and risks, but to support public debate in these regions. This approach aims to foster local ownership of human rights principles rather than external imposition.
Evidence
References specific EU projects in Africa and Latin America focused on promoting public debate
Major discussion point
Digital Divides and Connectivity Challenges
Topics
Development | Human rights
Agreed with
– Gitanjali Sah
– Dan York
– Participant
Agreed on
Digital divides and connectivity challenges must be addressed
Dan York
Speech speed
173 words per minute
Speech length
1154 words
Speech time
399 seconds
Technical communities bring essential expertise while civil society provides knowledge about impacts on vulnerable populations
Explanation
York argues that effective AI governance requires multi-stakeholder participation where different groups bring distinct value. Technical communities contribute essential knowledge about system design, while civil society provides crucial insights about how AI systems impact vulnerable populations in ways that may not be immediately apparent to technologists.
Evidence
Also mentions that governments bring policy information and end users provide crucial feedback about system impacts
Major discussion point
Multi-stakeholder Internet Governance and WSIS Plus 20
Topics
Human rights | Legal and regulatory
Agreed with
– Rasmus Lumi
– Gitanjali Sah
– Isabel Ebert
– Jan Gerlach
Agreed on
Multi-stakeholder approach is essential for effective Internet governance and AI governance
Disagreed with
– Isabel Ebert
Disagreed on
Scope and focus of multi-stakeholder engagement
Open standards and protocols are needed for AI transparency, explainability, and accountability
Explanation
York draws parallels between internet protocols like TCP/IP and HTTP that enabled global interoperability and the need for similar open standards for AI systems. He argues that standards for AI transparency, explainability, and accountability need to be developed at all layers of the AI stack through multi-stakeholder processes.
Evidence
References the Internet Engineering Task Force (IETF) as the standards organization that brought us foundational internet protocols
Major discussion point
AI Ethics and Human Rights Framework
Topics
Infrastructure | Legal and regulatory
Agreed with
– Isabel Ebert
– Jan Gerlach
Agreed on
Need for transparency, accountability and inclusion in AI governance
One-third of world (2.6 billion people) still lacks internet access, and AI development risks deepening digital divide
Explanation
York highlights that 2.6 billion people globally still lack internet access, and those who do have access don’t necessarily have affordable, reliable, or resilient connectivity. He warns that AI development is deepening the digital divide because those with access to AI tools become more productive while leaving offline populations further behind.
Evidence
Provides specific statistic of 2.6 billion people without internet access
Major discussion point
Digital Divides and Connectivity Challenges
Topics
Development | Digital access
Agreed with
– Gitanjali Sah
– Thibaut Kleiner
– Participant
Agreed on
Digital divides and connectivity challenges must be addressed
Those without connectivity cannot contribute knowledge to information pools used for training AI models
Explanation
York argues that the digital divide has implications beyond just access to AI tools – it also means that offline populations cannot contribute their knowledge and information to the pools of data being used to train AI models. This creates a feedback loop where AI systems lack diverse global perspectives.
Major discussion point
Digital Divides and Connectivity Challenges
Topics
Development | Human rights
Internet’s success built on open standards, open protocols, and innovation without permission principle
Explanation
York emphasizes that the internet’s success has been built on open standards, open protocols, and the principle that innovation can happen without permission. He contrasts this with earlier times when you couldn’t put anything online without paying somebody, arguing that this openness has been fundamental to internet development.
Evidence
References the Internet Society’s founding in 1992 and the IETF’s role in developing protocols like TCP/IP and HTTP
Major discussion point
Technical Infrastructure and Innovation
Topics
Infrastructure | Legal and regulatory
Concern about proprietary, closed AI systems creating vendor lock-in and concentrated power
Explanation
York expresses concern from the technical community about the trend toward proprietary, closed AI systems that create the same problems seen in parts of the internet today. He warns about vendor lock-in, closed proprietary systems, and concentrated power as risks that need to be addressed.
Major discussion point
Technical Infrastructure and Innovation
Topics
Economic | Legal and regulatory
Need to balance protecting against harms while ensuring continued innovation
Explanation
York acknowledges the challenge of balancing protection against AI harms with maintaining the internet’s tradition of innovation without permission. He argues for finding ways to protect against risks while ensuring that the level of innovation that has characterized internet development continues.
Major discussion point
Technical Infrastructure and Innovation
Topics
Legal and regulatory | Human rights
Internet Society chapters worldwide engage in connectivity and AI topics at national and regional levels
Explanation
York explains that the Internet Society has 120 chapters around the world that engage in these discussions at national and regional levels. While many focus primarily on connectivity issues, some are very involved with AI topics, and there are also about 180 national or regional Internet Governance Forums bringing people together globally.
Evidence
Provides specific numbers: 120 Internet Society chapters and 180 national/regional Internet Governance Forums
Major discussion point
Technical Infrastructure and Innovation
Topics
Development | Legal and regulatory
Ernst Noorman
Speech speed
133 words per minute
Speech length
1377 words
Speech time
619 seconds
Competition concerns should not override regulation designed to protect society and create level playing fields
Explanation
Noorman argues against calls to pause regulatory frameworks like the EU AI Act for competition reasons, emphasizing that regulation serves to protect society and often improves innovation by creating level playing fields. He warns against prioritizing competitive advantage over societal protection in AI governance.
Evidence
References recent calls in Europe to pause the EU AI Act due to competition concerns
Major discussion point
International Cooperation and Regulatory Approaches
Topics
Legal and regulatory | Economic
Disagreed with
– Thibaut Kleiner
Disagreed on
Approach to AI regulation and competition concerns
Participant
Speech speed
122 words per minute
Speech length
290 words
Speech time
142 seconds
Technical institutions like ISOC, IEEE, IGF need to enhance regional coordination and support national chapters to complement state member efforts
Explanation
The participant argues that while technical institutions like the Internet Society make highly valuable contributions to the Internet community, there is a need to enhance regional coordination offices and support national office chapters. They suggest this would help complement the efforts of state members and ITU in addressing connectivity challenges, noting that 2 billion people remain unconnected despite previous large-scale conferences.
Evidence
References ISOC Congress hosted in Geneva 2012 with 3,000 delegates, and mentions that 2 billion people still lack connectivity as stated by the Secretary General
Major discussion point
Technical Infrastructure and Innovation
Topics
Development | Infrastructure
Agreed with
– Gitanjali Sah
– Thibaut Kleiner
– Dan York
Agreed on
Digital divides and connectivity challenges must be addressed
Agreements
Agreement points
Multi-stakeholder approach is essential for effective Internet governance and AI governance
Speakers
– Rasmus Lumi
– Gitanjali Sah
– Isabel Ebert
– Jan Gerlach
– Dan York
Arguments
Multi-stakeholder approach must be inclusive, participatory, and consensus-driven involving public sector, private sector, civil society, technical community, and academia
WSIS Plus 20 provides opportunity for multi-stakeholder dialogue to include all voices in the UN General Assembly review process
Making rules inclusive requires expanding participation beyond states to include affected communities in AI governance decisions
Civil society input is critical for internet governance success, but their access to these processes is under threat
Technical communities bring essential expertise while civil society provides knowledge about impacts on vulnerable populations
Summary
All speakers strongly advocate for inclusive multi-stakeholder participation in Internet and AI governance, emphasizing that effective governance requires input from diverse actors including governments, private sector, civil society, technical community, and academia
Topics
Human rights | Legal and regulatory
Human rights framework should be central to digital governance and AI development
Speakers
– Rasmus Lumi
– Gitanjali Sah
– Isabel Ebert
– Thibaut Kleiner
Arguments
Freedom Online Coalition provides coordinated response through multi-stakeholder format
Universal Declaration of Human Rights featured prominently in original WSIS outcome documents
Human rights framework should serve as common minimum denominator for ethical approach to technology
European Union’s declaration on digital rights and principles puts people and rights at center of digital transformation
Summary
Speakers agree that human rights principles should be the foundational framework for digital governance, with rights-based approaches being essential for ethical technology development and deployment
Topics
Human rights | Legal and regulatory
Need for transparency, accountability and inclusion in AI governance
Speakers
– Isabel Ebert
– Jan Gerlach
– Dan York
Arguments
Need transparent rules matching pace of AI development with benefits shared across nations and risks thoughtfully managed
Blueprint for information integrity supports individual agency through literacy, privacy, safety and transparency
Open standards and protocols are needed for AI transparency, explainability, and accountability
Summary
Speakers emphasize the critical importance of building transparency, accountability mechanisms, and inclusive participation into AI governance frameworks to ensure responsible development and deployment
Topics
Human rights | Legal and regulatory
Digital divides and connectivity challenges must be addressed
Speakers
– Gitanjali Sah
– Thibaut Kleiner
– Dan York
– Participant
Arguments
Need more cross-sectoral work and best practices sharing across countries and stakeholders
EU supports projects engaging with Africa and Latin America to promote public debate about digital technology challenges
One-third of world (2.6 billion people) still lacks internet access, and AI development risks deepening digital divide
Technical institutions like ISOC, IEEE, IGF need to enhance regional coordination and support national chapters to complement state member efforts
Summary
Speakers recognize that significant portions of the global population remain unconnected and that digital divides risk being deepened by AI development, requiring coordinated efforts to bridge these gaps
Topics
Development | Digital access
Similar viewpoints
Both speakers emphasize the responsibility of private sector actors in ensuring ethical AI development, with Ebert focusing on companies building human rights into products and Gerlach highlighting the need to support communities that create the data used to train AI systems
Speakers
– Isabel Ebert
– Jan Gerlach
Arguments
Companies must build human rights into their products and services following UN Guiding Principles on Business and Human Rights
Good AI governance requires supporting communities who curate and verify information that feeds AI systems
Topics
Human rights | Legal and regulatory
Both speakers advocate for comprehensive, cross-sectoral approaches to AI governance that respect local contexts while maintaining universal principles
Speakers
– Gitanjali Sah
– Thibaut Kleiner
Arguments
AI governance must be cross-sectoral, looking across health, agriculture, education with overarching ethical framework
Human rights approach is universal and should be bottom-up, owned by local constituencies rather than imposed
Topics
Human rights | Legal and regulatory
Both speakers emphasize the foundational role of open, collaborative approaches in building Internet infrastructure, with York focusing on technical standards and Gerlach on civil society contributions
Speakers
– Dan York
– Jan Gerlach
Arguments
Internet’s success built on open standards, open protocols, and innovation without permission principle
Civil society builds large parts of internet including digital public goods that Global Digital Compact aims to support
Topics
Infrastructure | Development
Unexpected consensus
Resistance to centralized state control over Internet governance
Speakers
– Rasmus Lumi
– Dan York
– Jan Gerlach
Arguments
Need to resist attempts to overturn distributed multi-stakeholder Internet governance model and prevent expansion of state control
Concern about proprietary, closed AI systems creating vendor lock-in and concentrated power
Civil society input is critical for internet governance success, but their access to these processes is under threat
Explanation
Unexpected consensus emerged around concerns about centralization of power, whether by states or private actors, with speakers from different backgrounds (government, technical community, civil society) all expressing concerns about threats to distributed governance models
Topics
Legal and regulatory | Human rights
AI systems should serve society rather than drive technological determinism
Speakers
– Isabel Ebert
– Ernst Noorman
Arguments
Need to shift debate to asking what kind of societies we want AI to help build rather than just adapting to AI
Competition concerns should not override regulation designed to protect society and create level playing fields
Explanation
Unexpected alignment between human rights advocate and government representative on rejecting technological determinism and prioritizing societal needs over competitive or technological imperatives
Topics
Human rights | Legal and regulatory
Overall assessment
Summary
Strong consensus emerged around core principles of multi-stakeholder governance, human rights-centered approaches, and the need for inclusive, transparent AI governance. Speakers consistently emphasized the importance of maintaining distributed governance models while addressing digital divides and ensuring ethical technology development.
Consensus level
High level of consensus on fundamental principles with broad agreement across different stakeholder groups (government, international organizations, civil society, technical community) on the need for rights-based, inclusive approaches to digital governance. This strong alignment suggests potential for coordinated action in WSIS Plus 20 processes and beyond, though implementation challenges remain around balancing innovation with protection and ensuring meaningful participation of all stakeholders.
Differences
Different viewpoints
Approach to AI regulation and competition concerns
Speakers
– Ernst Noorman
– Thibaut Kleiner
Arguments
Competition concerns should not override regulation designed to protect society and create level playing fields
European Union’s declaration on digital rights and principles puts people and rights at center of digital transformation
Summary
While both support regulation, Noorman explicitly argues against pausing AI regulation for competition reasons, while Kleiner focuses on the EU’s regulatory approach without addressing the competition vs. regulation tension
Topics
Legal and regulatory | Economic
Scope and focus of multi-stakeholder engagement
Speakers
– Dan York
– Isabel Ebert
Arguments
Technical communities bring essential expertise while civil society provides knowledge about impacts on vulnerable populations
Making rules inclusive requires expanding participation beyond states to include affected communities in AI governance decisions
Summary
York emphasizes the distinct roles of technical communities and civil society, while Ebert focuses more broadly on expanding participation to affected communities beyond traditional stakeholders
Topics
Human rights | Legal and regulatory
Unexpected differences
Institutionalization vs. flexibility in governance structures
Speakers
– Thibaut Kleiner
– Dan York
Arguments
Internet Governance Forum should become permanent UN institution with own budget and director for ongoing discussions about emerging technologies
Need to balance protecting against harms while ensuring continued innovation
Explanation
Kleiner advocates for formal institutionalization of the IGF, while York emphasizes maintaining the internet’s tradition of ‘innovation without permission’ and warns against proliferation of formal mechanisms that could hinder participation
Topics
Legal and regulatory | Infrastructure
Overall assessment
Summary
The discussion shows remarkable consensus on core principles (multi-stakeholder governance, human rights protection, AI ethics) but reveals nuanced differences in implementation approaches and priorities
Disagreement level
Low to moderate disagreement level. Most differences are complementary rather than contradictory, focusing on different aspects of the same challenges. The main tensions arise around balancing formal regulation with innovation flexibility, and different emphases on technical vs. social approaches to governance. These disagreements reflect healthy diversity in problem-solving approaches rather than fundamental conflicts, which could strengthen overall policy development if properly integrated.
Partial agreements
Partial agreements
Similar viewpoints
Both speakers emphasize the responsibility of private sector actors in ensuring ethical AI development, with Ebert focusing on companies building human rights into products and Gerlach highlighting the need to support communities that create the data used to train AI systems
Speakers
– Isabel Ebert
– Jan Gerlach
Arguments
Companies must build human rights into their products and services following UN Guiding Principles on Business and Human Rights
Good AI governance requires supporting communities who curate and verify information that feeds AI systems
Topics
Human rights | Legal and regulatory
Both speakers advocate for comprehensive, cross-sectoral approaches to AI governance that respect local contexts while maintaining universal principles
Speakers
– Gitanjali Sah
– Thibaut Kleiner
Arguments
AI governance must be cross-sectoral, looking across health, agriculture, education with overarching ethical framework
Human rights approach is universal and should be bottom-up, owned by local constituencies rather than imposed
Topics
Human rights | Legal and regulatory
Both speakers emphasize the foundational role of open, collaborative approaches in building Internet infrastructure, with York focusing on technical standards and Gerlach on civil society contributions
Speakers
– Dan York
– Jan Gerlach
Arguments
Internet’s success built on open standards, open protocols, and innovation without permission principle
Civil society builds large parts of internet including digital public goods that Global Digital Compact aims to support
Topics
Infrastructure | Development
Takeaways
Key takeaways
Multi-stakeholder Internet governance model is under threat and must be defended against attempts to expand state control, particularly in the WSIS Plus 20 review process
AI governance requires a cross-sectoral, human rights-based approach with transparency, accountability, and inclusion as core principles
The digital divide is deepening as 2.6 billion people remain unconnected while AI development accelerates, risking further marginalization of offline populations
Civil society participation in Internet governance processes is critical but increasingly threatened, requiring protection and enhancement of their access
Open standards and protocols are essential for AI systems to ensure transparency, explainability, and accountability, avoiding proprietary lock-in
Human rights framework should serve as the common minimum denominator for ethical technology development, with the UN Guiding Principles on Business and Human Rights providing guidance for corporate responsibility
Information integrity requires supporting communities that curate and verify information used to train AI systems, including projects like Wikipedia
Regional and international cooperation must focus on bottom-up, locally-owned approaches to human rights rather than top-down imposition
Resolutions and action items
Freedom Online Coalition member states to work together to take advisory network feedback into account in their national positions for WSIS Plus 20
Make the Internet Governance Forum a permanent UN institution with its own budget and director
Develop concrete recommendations to complement both the WSIS Plus 20 process and FOC’s ongoing efforts to uphold Internet freedom
Support civil society organizations through smart policies and funding to sustain trustworthy information ecosystems
Ensure multi-stakeholder involvement is maintained in the Open-Ended Working Group on responsible state behavior in ICTs
Continue providing platforms for multi-stakeholder dialogue through WSIS Forum and IGF processes with remote participation and regional accommodation
Unresolved issues
How to effectively balance different voices within the multi-stakeholder model and define what constitutes meaningful multi-stakeholder participation
Whether current international human rights frameworks are adequately equipped to respond to rapid AI evolution
How to prevent AI development from deepening existing inequalities while maintaining innovation momentum
How to connect the remaining 2.6 billion unconnected people and ensure their knowledge contributes to AI training data
How to balance protecting against AI harms while preserving the ‘innovation without permission’ principle that enabled Internet success
How to resist calls to pause AI regulation for competition reasons while maintaining technological leadership
How to ensure meaningful participation of affected communities who cannot physically attend governance forums
How to develop AI standards for transparency and accountability across all layers of the AI stack
Suggested compromises
Integrate multi-stakeholder involvement into multilateral forums rather than replacing multilateral approaches entirely
Use incentive-based stimulus packages alongside regulation to encourage responsible AI development
Focus on creating dialogue and promoting discussion with international partners rather than imposing human rights views from outside
Accommodate civil society participation through remote access and regional time zone considerations when physical attendance is not possible
Leverage existing national and regional Internet Governance Forums (180 worldwide) as venues for broader participation rather than creating new mechanisms
Build on existing WSIS framework and action lines that already incorporate human rights principles rather than starting from scratch
Thought provoking comments
We cannot overcome the challenges without the meaningful engagement of all shareholders. We need coordinated response… We must adopt a strong common approach to ensure the protection of the Internet’s decentralized model. Efforts to impose centralized control threaten to undermine the Internet’s fundamental openness, risking fragmentation and compromising the very attributes that have made the Internet a catalyst for progress and innovation.
Speaker
Rasmus Lumi
Reason
This comment was insightful because it framed the entire discussion around a fundamental tension between centralized state control versus decentralized multi-stakeholder governance. It established the stakes of the conversation – that the very nature of internet governance is under threat and requires active defense.
Impact
This opening comment set the tone for the entire discussion, with subsequent speakers repeatedly returning to themes of multi-stakeholder engagement, the importance of inclusive governance, and resistance to state control. It created a sense of urgency that permeated all following contributions.
Instead of asking how do we adapt AI, we should ask what kind of societies do we want AI to help us build and which accountability structures for different actors and their distinct role can incentivise this.
Speaker
Isabel Ebert
Reason
This comment was particularly thought-provoking because it fundamentally reframed the AI governance debate from a reactive to a proactive stance. Rather than adapting to AI’s development, it suggested we should first define our societal goals and then shape AI to serve those purposes.
Impact
This reframing influenced subsequent speakers to focus more on human-centered approaches. It shifted the discussion from technical adaptation to values-based design, with later speakers like Jan Gerlach emphasizing that ‘the internet really must serve’ people and Dan York discussing ‘innovation without permission’ as a core principle.
Civil society, the people who use the internet, also build large parts of the internet… They build the digital public goods that the Global Digital Compact aims to support… Take Wikipedia as an example… To ensure such projects can continue to thrive, these people, through civil society organizations, need to have a voice at the table of Internet governance processes.
Speaker
Jan Gerlach
Reason
This comment was insightful because it challenged the typical framing of civil society as merely users or beneficiaries of technology, instead positioning them as active builders and creators of internet infrastructure. It highlighted an often-overlooked contribution of civil society to the digital ecosystem.
Impact
This comment deepened the discussion by adding a new dimension to multi-stakeholder engagement – not just consultation but recognition of civil society as infrastructure builders. It influenced the moderator’s later observation about balancing different voices in the multi-stakeholder model and added weight to arguments for meaningful civil society participation.
A third of the world is still not connected. There’s 2.6 billion people who do not have access to the internet… We’re, in fact, with some of what we’re doing, we are deepening the digital divide. Because those of us who have access to the AI tools and systems that we’re all using, we are able to be more productive… And we’re leaving the folks who are offline further behind.
Speaker
Dan York
Reason
This comment was particularly thought-provoking because it introduced a sobering reality check about digital inequality in the context of AI advancement. It highlighted how technological progress can paradoxically worsen existing inequalities rather than solve them.
Impact
This comment brought a critical equity lens to the discussion that had been somewhat abstract until this point. It grounded the conversation in concrete numbers and consequences, influencing the moderator’s concluding remarks about ensuring AI serves society and humanity rather than just competition.
Right now, if you look at the AI discussion, it’s more and more about competition. Who will be the winner?… AI is there to serve society and humanity. And how do we ensure that?… Just these days, you can hear even in Europe, calls to pause the EU AI Act because of competition reasons. I think that’s not the good way to go.
Speaker
Ernst Noorman (Moderator)
Reason
This concluding comment was insightful because it crystallized a key tension that had been building throughout the discussion – the conflict between competitive economic interests and societal protection. It directly challenged the prevailing narrative that regulation hinders innovation.
Impact
As a concluding comment, this synthesized many of the discussion’s themes and provided a clear call to action. It reinforced the human rights-centered approach advocated by earlier speakers and positioned regulation as an enabler rather than inhibitor of beneficial innovation.
Overall assessment
These key comments shaped the discussion by establishing a clear narrative arc from problem identification to solution frameworks. Lumi’s opening created urgency around defending multi-stakeholder governance, Ebert’s reframing shifted focus to proactive, values-based AI development, Gerlach’s contribution elevated civil society from beneficiaries to builders, York’s reality check grounded the discussion in equity concerns, and Noorman’s conclusion synthesized these themes into a call for human-centered rather than competition-driven approaches. Together, these comments transformed what could have been a technical policy discussion into a more fundamental conversation about power, values, and the future of digital governance. The progression showed how individual insights can build upon each other to deepen collective understanding and create momentum for action.
Follow-up questions
How can we work together to prevent AI technologies from deepening existing inequalities?
Speaker
Ernst Noorman
Explanation
This was identified as a key area to explore in the session introduction, focusing on ensuring AI doesn’t exacerbate current social and economic disparities
Are current international and human rights frameworks equipped to respond to the rapid evolution of AI?
Speaker
Ernst Noorman
Explanation
This question addresses whether existing legal and regulatory frameworks are adequate for governing rapidly advancing AI technologies
How do children know about their rights online, and are schools educating them properly?
Speaker
Gitanjali Sah
Explanation
This highlights the need for better digital literacy and rights education for children in the context of online safety and governance
Do we have the right governance structures and guidelines for parents and educators regarding children’s online rights?
Speaker
Gitanjali Sah
Explanation
This identifies a gap in support systems for adults responsible for children’s digital wellbeing and education
What kind of societies do we want AI to help us build, and which accountability structures can incentivize this?
Speaker
Isabel Ebert
Explanation
This reframes the AI governance debate from technical adaptation to societal vision and appropriate accountability mechanisms
How can we enhance regional coordination offices and support national office chapters of organizations like ISOC?
Speaker
Riyad Abathia
Explanation
This addresses the need for stronger local and regional representation in internet governance processes, particularly for civil society participation
How do we balance different voices in the multistakeholder model?
Speaker
Ernst Noorman
Explanation
This fundamental question about multistakeholder governance seeks to understand how to ensure equitable representation and influence among different stakeholder groups
How do we ensure AI serves society and humanity rather than just competition and economic interests?
Speaker
Ernst Noorman
Explanation
This addresses the tension between commercial AI development focused on competition versus AI development that prioritizes societal benefit
What standards can be developed for AI transparency, explainability, and accountability across all layers of the AI stack?
Speaker
Dan York
Explanation
This identifies the need for comprehensive technical standards that ensure AI systems are transparent and accountable at every level of their operation
How can we connect the 2.6 billion people who still lack internet access to prevent deepening digital divides in the AI era?
Speaker
Dan York
Explanation
This highlights the urgent need to address basic connectivity issues to prevent AI from further marginalizing already disconnected populations
Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.