This session, organised by ICC BASIS, brought together representatives from business, the United Nations, and government to discuss how the private sector can contribute to implementing the outcomes of the WSIS+20 review and advance broader digital cooperation efforts, including those related to the Global Digital Compact .
Helen Harris outlined some concrete contributions by Amazon across several WSIS action lines, including the deployment of the Amazon Kuiper low Earth orbit satellite constellation to bridge the digital divide , the provision of free cloud and AI skills training to over 31 million people across more than 200 countries , and the operation of disaster relief hubs that have delivered over 26 million relief items across 200-plus disasters since 2017 . Meta's Rachel Silley highlighted the company's investment in subsea cable infrastructure through the ITU's Partners2Connect Coalition, its open-source AI tools being applied in sectors such as agriculture and healthcare, and its Small Business Growth Academy launched across 12 Asia-Pacific countries .
Gitanjali Sah, from the ITU, described the UN Group on the Information Society (UNGIS) as a key coordination mechanism, noting that member states had instructed it to engage more actively with non-governmental stakeholders . UNGIS received approximately 200 nominations for a new multi-stakeholder consultation group and is also establishing a digital finance task force involving organisations such as the OECD, World Bank, and UNCTAD .
UNDP's Yu Ping Chan welcomed the multi-stakeholder approach but cautioned that conversations must translate into concrete guidance and accountability for both governments and the private sector . She cited a practical example of UNDP brokering GPU access and cloud credits for African startups through partnerships with the Italian government, Microsoft, and AWS . Sebastian Mhatre from the UK's FCDO emphasised three roles for the private sector: transparency around local value capture, collaboration on interoperability, and support for digital exports from developing countries .
A question from the audience raised the issue of the compute divide facing the Global South, with panellists suggesting that blended public-private partnerships, open-source models, and frugal or small-scale AI approaches could offer pragmatic pathways forward . Overall, the discussion underscored that sustained, concrete multi-stakeholder collaboration - rather than high-level declarations alone - is essential to translating WSIS commitments into meaningful outcomes on the ground .
Overall Purpose
- The discussion is an ICC (International Chamber of Commerce) workshop focused on business perspectives for advancing the outcomes of the WSIS+20 review. The session aims to highlight the private sector's role in implementing WSIS action lines, explore how businesses can engage with the UNGIS group and UN partners, and translate WSIS commitments into concrete, scalable initiatives through sustained multi-stakeholder collaboration. ---
Major Discussion Points
- Private sector contributions to WSIS action lines through concrete initiatives: Several panellists shared tangible examples of business implementation across WSIS action lines. Amazon highlighted its Project Kuiper satellite broadband constellation to bridge the digital divide , free cloud and AI skills training provided to over 31 million people across 200+ countries , and disaster relief operations delivering over 26 million relief items across 200+ disasters since 2017 . Meta described investments in subsea cable infrastructure through the ITU's Partners2Connect Coalition, its Workforce Academy for skilled trades, and open-source AI tools being applied in agriculture and healthcare .
- The role of UNGIS in enabling multi-stakeholder coordination: Gitanjali Sah outlined how UNGIS, as the UN's interagency coordination mechanism for digital cooperation, is evolving to engage more directly with non-governmental stakeholders. A new multi-stakeholder consultation group received approximately 200 nominations, reflecting strong interest from both experienced and new participants . A digital finance task force is also being established to explore gaps in digital financing for developing countries, with commitments to consult the private sector as part of its terms of reference .
- The need to translate WSIS commitments into concrete, actionable outcomes: Yu Ping Chan of UNDP raised a critical concern about the gap between high-level commitments and on-the-ground implementation. She questioned whether the multi-stakeholder conversations around WSIS+20 and the Global Digital Compact had yet produced meaningful guidance for businesses, national governments, and civil society . She called for concrete outcomes that could change behaviour and conditions in developing countries, rather than remaining at the level of declarations and talking points .
- Financing for digital development and the role of blended public-private investment: The session addressed the WSIS+20 call for a task force on financial mechanisms for digital development. Rachel Silley of Meta emphasised that private capital deployment in backbone infrastructure, such as subsea cables, operates at a scale that public financing alone cannot match, and cited this as a concrete example of the blended cooperation the WSIS+20 outcome document calls for . Sebastian Mhatre identified three areas for private sector contribution: transparency around local value capture, interoperability, and supporting local businesses in developing countries to participate in digital exports .
- Bridging the compute divide in the Global South: An audience question raised concerns about the gap in compute resources between advanced economies and developing countries, and how the Global South could participate in the AI production ecosystem . Panellists offered several responses: UNDP described a partnership providing GPU access and cloud credits to African startups through the AI Help for Sustainable Development initiative . Both Sebastian Mhatre and Yu Ping Chan advocated for exploring low-compute, edge-compute, and small language model approaches tailored to specific development use cases as a pragmatic alternative to replicating large-scale infrastructure . Rachel Silley also pointed to open-source models as a means for local developers to adapt AI to their domestic contexts .
- --
Overall Tone
- The overall tone of the discussion is collaborative, constructive, and professionally optimistic. Panellists consistently emphasised partnership, shared goals, and the importance of multi-stakeholder engagement, reflecting a spirit of goodwill between the UN, business, and government representatives.
- The tone shifted when Yu Ping Chan introduced a more candid and critical perspective, underscoring the need to translate commitments into concrete action . This moment introduced a degree of productive tension into the conversation.
- Towards the end of the session, the tone returned to solution-oriented and forward-looking, particularly during the discussion on compute access in the Global South, where panellists offered practical suggestions and concrete examples rather than dwelling on the scale of the challenge. The closing remarks reinforced a collegial and encouraging atmosphere, with acknowledgement of the business community's contributions and an invitation to continue engagement at forthcoming events.
ICC BASIS Workshop: Business Perspectives on Advancing WSIS Plus 20 Outcomes
#
Session Overview and Purpose
The ICC BASIS Workshop brought together representatives from business, the United Nations, and government to examine how the private sector can contribute to implementing the outcomes of the WSIS Plus 20 review and advance broader digital cooperation efforts . The session was moderated by Meni Anastasiadou, Digital Policy Manager at the International Chamber of Commerce, who stepped in for her colleague Tamash Udo, who was expected to join later due to scheduling commitments during a busy week . The session was structured around three interconnected themes: the role of business across the WSIS architecture; the enabling policy environment for digital transformation; and financing mechanisms for digital development . The overarching goal was to move beyond high-level declarations and identify concrete, scalable initiatives through sustained multi-stakeholder collaboration, including those related to the Global Digital Compact .
The panel comprised Helen Harris, Manager of Public Policy at Amazon; Rachel Silley from Meta; Yuping Chan, Head of Digital Partnerships and Engagement at UNDP; Gitanjali Sah, Head of the WSIS and Stakeholder Engagement Division at the ITU; and Sebastian Mhatre, Senior Digital Development Advisor at the UK's Foreign, Commonwealth and Development Office . The discussion was structured to present the role of business in contributing to the WSIS Plus 20 follow-up process, since the outcome document was adopted in December of the previous year .
---
#
Private Sector Contributions to WSIS Action Lines
Helen Harris opened the substantive discussion by outlining Amazon's concrete contributions across several WSIS action lines, framing them as direct implementations of the WSIS framework rather than standalone corporate initiatives . On Action Line C2, concerning information and communication infrastructure, she described Amazon Leo - a low Earth orbit satellite broadband constellation of 3,236 planned satellites - as a mechanism for bringing affordable high-speed connectivity to unserved and underserved communities globally, directly advancing the WSIS goal of bridging the digital divide .
On Action Line C4, capacity building, Harris reported that Amazon had provided free cloud and AI skills training to over 31 million people across more than 200 countries and territories, surpassing its original target of 29 million people by 2025 a full year ahead of schedule . The company's Skills-to-Jobs Tech Alliance, in over a dozen countries and encompassing more than 990 educational institutions and 700 employers across the UK, Singapore, India, Colombia, Brazil, Malaysia, and the United States, was cited as a vehicle specifically designed to connect learners to employers .
On Action Line C11, international and regional cooperation, Harris highlighted Amazon's disaster relief hubs, which partner directly with UN agencies including IOM and WFP, as well as the Red Cross and Save the Children . Since 2017, Amazon has delivered more than 26 million relief items across more than 200 disasters worldwide through in-kind donations and pro bono logistics . A recent example involved the earthquakes in Venezuela, where Amazon activated its network, donated directly to IOM and WFP to help coordinate emergency supplies, and donated over 660,000 emergency supplies as transportation routes reopened . Harris concluded by arguing that underpinning all of these contributions is Action Line C1 - the governance approach itself - and that none of it happens without the multi-stakeholder table .
Rachel Silley of Meta reinforced this picture of private sector engagement, noting that Meta has been an partner in the global governance and implementation of WSIS since its inception . She emphasised that the accelerating pace of digital transformation makes discussions on financing even more urgent , and that the technology sector brings unique real-time expertise in innovation, infrastructure development, deployment, and technical standards . Meta's specific contributions included a commitment to the ITU's Partners2Connect Coalition focused on bridging the digital divide through the build-out of robust subsea cable infrastructure - described as the kind of private capital deployment that public financing alone cannot match at scale . Silley also described Meta's Workforce Academy, recently launched in the United States to create accessible pathways into skilled trades such as electricians, mechanics, and fibre technicians needed to support AI adoption . On AI accessibility, she highlighted Meta's open-source LLAMA model being used to identify potential reducers of crop yields and the Mayo Clinic's use of LLAMA2 to improve radiation therapy decision-making , as well as the Small Business Growth Academy launched across 12 Asia-Pacific countries to help small businesses use digital and AI-enabled tools .
---
#
UNGIS as a Coordination Mechanism and the Multi-Stakeholder Consultation Group
Gitanjali Sah of the ITU opened her remarks by acknowledging that ICC has been the voice of the private sector in the WSIS process since its inception . She then provided historical context for the WSIS process itself, noting that it was conceived at the 1997 Minneapolis ITU Plenipotentiary Conference, proposed by Tunisia as a member state, and subsequently taken to the United Nations, which proposed a two-phased process - one summit in Tunisia and one in Geneva . She noted that the process was inherently multi-stakeholder from the outset, with the recognition that a process looking at technologies could not proceed without the private sector, civil society, and the technical community . She acknowledged the presence of stalwarts in the process, including Ambassador Karklings and Professor Winkin, whose consistent dedication had contributed to what had been achieved .
Turning to the post-WSIS Plus 20 landscape, Sah described the UN Group on the Information Society (UNGIS) as an interagency mechanism created to coordinate discussions within the United Nations system, currently encompassing 50 UN agencies on digital cooperation . She noted the high-level commitment to UNGIS, evidenced by the Secretary General-hosted UNGIS high-level lunch attended by 10 heads of UN agencies, at least 15 deputies, and around 50 UN agencies . ITU has taken over as chair and serves as the permanent secretariat of UNGIS .
A significant development described by Sah was the launch of a new Multi-Stakeholder Consultation Group (MCG) within UNGIS, following a mandate from member states to engage more actively with non-governmental stakeholders . The nomination process, conducted in consultation with UNDP, UNESCO, UNCTAD, and UNECE, yielded approximately 200 nominations - far exceeding expectations - from both experienced practitioners and new participants in the WSIS process . Sah expressed enthusiasm about the diversity of interest, noting that the WSIS Forum reflected a balance between those who had maintained the process and new people wanting to be part of it . One model under consideration involves a core group at the international level complemented by regional and national representations . However, she was clear that the MCG would be a consultative group and would not interfere in the day-to-day business of UNGIS . Sah also referenced the second Business Roundtable during her remarks, crediting ICC, USCIB, and others as co-organisers .
A second institutional development was the creation of a digital finance task force under UNGIS, drawing on experts from OECD, World Bank, UNCTAD, ITU, UNESCO, and UNDP, with a mandate to explore gaps in digital finance in developing countries . Sah confirmed that the private sector, as well as multilateral and regional banks, would be consulted as part of the task force's terms of reference .
Helen Harris had earlier made a structural recommendation that directly informed this discussion, calling for UNGIS to function as a single entry point for business engagement with the UN system - the single door business knocks on instead of a dozen separate ones - and arguing that UNGIS can encourage and strengthen collaboration with industry so that business input feeds more directly into the work the action line facilitators are already doing .
---
#
The Enabling Policy Environment and the Challenge of Translating Commitments into Action
Yuping Chan of UNDP offered a broader perspective on the enabling policy environment, noting that UNDP is present in 170 countries and supports over 130 governments in using digital and AI to achieve the SDGs . She observed that the acceleration in interest around digital and AI came largely after COVID, when there was a global realisation that digital technologies are essential to achieving development outcomes . When WSIS was first conceptualised 20 years ago, notions around enabling environments were considered more nascent and fringe; today, with digital technologies and AI ubiquitous across almost every part of society, the need for government and public sector intervention to harness these technologies has become even more important . This is particularly critical in developing countries, which are making foundational choices about digital and AI adoption, often for the first time, and building digital ecosystems across entire societies and communities .
However, Chan introduced a notably candid and critical perspective that shifted the tone of the discussion. She questioned whether the multi-stakeholder conversations around WSIS Plus 20 and the Global Digital Compact had yet produced anything more meaningful than conversations, expressing concern about a tendency towards what she described as "UN talking point bingo where a couple of meetings become all that is" . She described struggling personally with how to translate words on paper into concrete guidance for both private sector and national governments, and questioned whether the processes had reached the point of actually landing in action that guides what businesses and governments do . She called explicitly for concrete outcomes and guidance - beyond the MCG and conversations - that would specify what is expected of private sector, national governments, and civil society, and how accountability would be maintained . As someone focused on in-country implementation, she emphasised UNDP's concern with what concrete things can be drawn out of these processes and given to national governments and multinational corporations to actually change behaviour and conditions on the ground .
This contribution introduced a note of candour and critical reflection into what had otherwise been a largely constructive exchange of achievements and frameworks. It implicitly challenged other panellists to move beyond listing initiatives and towards addressing the accountability gap between high-level commitments and on-the-ground impact.
---
#
Financing for Digital Development and the Private Sector's Role
The session addressed the WSIS Plus 20 call for a task force on financial mechanisms for digital development, with the moderator noting that this leads to questions about what role the private sector could eventually play . Rachel Silley argued that multi-stakeholder collaboration has always been key to the success of WSIS and is particularly relevant with financing for development, with the technology sector bringing unique real-time expertise that makes inclusive decision-making essential for robust, practical, and globally accepted solutions . She presented Meta's subsea cable infrastructure investment as a concrete example of the blended public-private cooperation the WSIS Plus 20 outcome document explicitly calls for, noting that private capital deployment at this scale cannot be matched by public financing alone .
Sebastian Mhatre of the UK FCDO offered remarks on private sector contribution to digital development financing, though the transcript captures only a portion of his contribution, beginning mid-speech. As recorded, his remarks covered three areas: the importance of local value capture in the digital economy, noting that most developing countries currently export raw materials or add only small amounts of value, and that digital technology offers an opportunity to compete at a higher level on the value chain ; the importance of working with the multi-stakeholder community on interoperability ; and the potential for private sector partnerships with local businesses in developing countries to support digital exports, creating opportunities for countries to participate in a new area of the global economy where they had previously been absent .
---
#
Bridging the Compute Divide in the Global South
An audience question from Jimson Olufouye of Afikta raised a concern that had not been addressed by the main panel: the gap in compute resources and infrastructure between developed and developing countries, and how developing countries could be included in the AI production ecosystem rather than merely as consumers . Due to technical difficulties, Jimson's question was relayed and paraphrased by another audience member, Amir, before panellists responded.
Sebastian Mhatre acknowledged the compute divide as a massive divide in terms of investment, noting that even advanced economies struggle to mobilise the required capital, making it very difficult to see how the gap is bridged through conventional means . Rather than dismissing the challenge, he proposed exploring alternative approaches including low compute options, edge compute options, and locally run small language models, arguing that these can create opportunities for a different type of innovation ecosystem that is less dependent on the same scale of compute as advanced economies .
Yuping Chan agreed with this framing, pointing to frugal AI, smart AI, and small AI as pragmatic approaches that look at specific use cases in the Global South and ask what amount of GPU resource is actually reasonable and tailored to that specific use case . She also provided a concrete example of UNDP brokering public-private partnerships to support compute access: through the AI Help for Sustainable Development initiative under the Italian G7 presidency, Seneca - the Italian supercomputer - announced approximately 1.5 million GPU hours for 130 African startups, while Microsoft and AWS provided approximately three million dollars in cloud credits, already being deployed across 50 startups in sectors including health, agriculture, energy, and multilingual AI languages across 14 African countries outlined by Italy's Mate plan . She presented this as a replicable model for giving the Global South access to AI compute resources .
Rachel Silley added a further dimension, pointing to open-source AI models as a means for developers in developing countries to take existing models, tailor them to their specific needs, and build locally relevant applications without relying on large proprietary models that require significant compute resources . An Amazon/AWS representative - whose attribution in the transcript is noted with some ambiguity in the speaker labelling - noted that Amazon has made significant investments in capacity building and data centre deployment in the Global South, including in India and Brazil, and is working with governments on e-government and AI initiatives .
---
#
Environmental Sustainability of AI
Towards the close of the session, an audience member raised a concern via the chat that had not been addressed by the main panel: the environmental costs of AI, including its growing carbon footprint from enormous computing power and energy consumption . The question asked how businesses can balance innovation and profitability with the responsibility to build AI systems that are energy-efficient, climate-conscious, and aligned with global sustainability goals . The moderator acknowledged the point and noted it had been touched upon in earlier examples, though the session did not allow time for a substantive response from panellists .
---
#
Audience Contributions and Forward-Looking Commitments
Jan Skarklin, Head of Government Engagement at ICANN, used the Q&A to place on record the technical community's gratitude to the business community for their engagement in the WSIS Plus 20 review process, noting that without business participation the results might have been very different . Skarklin also identified himself as having been involved in the WSIS process from the very beginning, lending particular weight to his observations. He noted that the composition of the panel itself reflected a significant shift from 20 years ago, when the podium would have been occupied solely by grey-haired men . Looking ahead, he encouraged the business community to participate in the preparations for and proceedings of the Internet Governance Forum in Nairobi with full strength .
---
#
Closing Remarks and Next Steps
The session concluded with the moderator summarising several concrete next steps. UNGIS will convene after the WSIS Business Forum to explore how best to make use of the approximately 200 nominations received for the MCG, with a model under consideration involving a core international group complemented by regional and national representations . The digital finance task force will begin its substantive work once the WSIS Forum concludes, with a commitment to consult the private sector and multilateral and regional banks . A second Business Roundtable, co-organised by ICC, USCIB, and others, was announced for the following morning at 07:30 to continue discussions on private sector engagement in the WSIS follow-up process . A summary of the session's discussions was to be circulated to participants .
Overall, the discussion underscored a high degree of consensus across institutional boundaries on the indispensability of multi-stakeholderism, the concrete value of private sector contributions to WSIS action lines, and the necessity of blended public-private financing models. At the same time, Yuping Chan's candid challenge about the gap between multi-stakeholder rhetoric and on-the-ground impact served as an important counterweight, reinforcing that the real test of the WSIS Plus 20 outcomes lies not in the frameworks and consultation groups established, but in whether they produce measurable changes in behaviour and conditions for people in developing countries .
Private sector contributions to WSIS action lines through infrastructure, capacity building, and disaster relief - Amazon has deployed satellite broadband (Project Kuiper), trained 31 million people in cloud and AI skills, and delivered 26 million relief items across 200+ disasters
Arg. 1Helen Harris outlined Amazon's concrete contributions across multiple WSIS action lines, demonstrating that private sector engagement translates into tangible outcomes. These contributions span connectivity infrastructure, skills training, and humanitarian logistics, each aligned with specific WSIS action lines.
Amazon's Project Kuiper low Earth orbit satellite constellation of 3,236 planned satellites is being deployed to bring affordable high-speed connectivity to unserved and underserved communities globally . Free cloud and AI skills training has been provided to over 31 million people across more than 200 countries and territories, surpassing the original target of 29 million people by 2025 a full year ahead of schedule . Amazon's disaster relief hubs have delivered more than 26 million relief items across 200-plus disasters worldwide since 2017, including a recent response to earthquakes in Venezuela .
on: Private sector investment in connectivity infrastructure and capacity building directly advances WSIS goals and bridges the digital divide
on: Whether multi-stakeholder conversations under WSIS and the Global Digital Compact have translated into meaningful, concrete action or remain largely rhetorical
Business community's participation in the WSIS Plus 20 review process was instrumental in preserving the consensus-based, multi-stakeholder approach to internet governance
Arg. 2Helen Harris argued that the business community played a critical role in the WSIS Plus 20 review by working alongside like-minded governments to maintain the open, interconnected nature of the internet. This multi-stakeholder approach underpins all other digital development efforts.
During the WSIS Plus 20 review, Amazon worked alongside the broader business community and like-minded governments to help preserve the consensus-based approach that has kept the internet open, interconnected, and globally connected . Helen Harris noted that none of the concrete contributions happen without the multi-stakeholder table, framing governance (Action Line C1) as the foundation for all other action lines .
on: Multi-stakeholderism is the essential foundation of the WSIS process and must be preserved
UNGIS should function as a single entry point for business engagement with the UN system, reducing the burden of navigating multiple separate agencies
Arg. 3Helen Harris argued that UNGIS is uniquely positioned to serve as a centralised coordination point for business engagement with the UN, rather than requiring companies to approach each agency individually. This would make collaboration more efficient and ensure business input feeds directly into the work of action line facilitators.
Helen Harris described UNGIS as the UN's own interagency mechanism for coordinating WSIS and Information Society work, and suggested it could be the single door business knocks on instead of a dozen separate ones . She further noted that UNGIS can encourage and strengthen collaboration with industry so that business input feeds more directly into the work the action line facilitators are already doing .
on: Public-private partnerships and blended financing models are necessary to address the scale of digital development investment needs
Meta's investments in subsea cable infrastructure, workforce skills academies, and open-source AI tools directly advance WSIS goals around connectivity, capacity building, and inclusive digital transformation
Arg. 1Rachel Silley highlighted Meta's multi-pronged approach to advancing WSIS goals, encompassing physical infrastructure, workforce development, and accessible AI tools. These investments are framed as concrete examples of blended public-private cooperation called for in the WSIS Plus 20 outcome document.
Meta joined the ITU's Partners to Connect Coalition with a commitment to bridging the digital divide through the build-out of robust subsea cable infrastructure . Meta's Workforce Academy was recently launched in the United States as a fast-track programme for skilled trades careers, powered by a multi-million first-year investment . Meta's open, accessible AI tools, including the LLAMA model, are being applied across education, agriculture, and government efficiency, with examples including use by the Mayo Clinic for radiation therapy decision-making and crop yield identification for growers .
on: Private sector investment in connectivity infrastructure and capacity building directly advances WSIS goals and bridges the digital divide
on: Whether multi-stakeholder conversations under WSIS and the Global Digital Compact have translated into meaningful, concrete action or remain largely rhetorical
Private capital deployment in backbone infrastructure, such as subsea cables, complements public financing and represents the kind of blended public-private cooperation explicitly called for in the WSIS Plus 20 outcome document
Arg. 2Rachel Silley argued that private sector investment in critical connectivity infrastructure fills a gap that public financing alone cannot address at scale. This blended financing model is directly aligned with the WSIS Plus 20 outcome document's calls for public-private cooperation.
Meta's commitment to building out robust subsea cable infrastructure through the ITU's Partners to Connect Coalition was cited as an example of private capital deployment that public financing alone cannot match at scale . Rachel Silley explicitly noted this as a concrete example of the blended public-private cooperation that the WSIS Plus 20 outcome document calls for .
on: Public-private partnerships and blended financing models are necessary to address the scale of digital development investment needs
The private sector brings unique real-time expertise in innovation, infrastructure, and technical standards, making its inclusion in financing discussions essential for robust and practical solutions
Arg. 3Rachel Silley contended that the technology sector's expertise in innovation, infrastructure development, and technical standards is indispensable in financing discussions. Inclusive decision-making that incorporates the private sector leads to more robust, practical, and globally accepted solutions.
Rachel Silley stated that the technology sector brings unique real-time expertise in innovation, infrastructure development, deployment, and technical standards . She argued that engagement with governments, civil society, the technical community, and academia is not a courtesy but because inclusive decision-making leads to robust, practical, and globally accepted solutions .
on: Multi-stakeholderism is the essential foundation of the WSIS process and must be preserved
Open-source AI models allow developers in developing countries to tailor existing models to their local context, reducing dependency on large-scale compute infrastructure
Arg. 4Rachel Silley suggested that open-source AI models offer an alternative pathway for developers in the Global South to access and adapt AI capabilities without requiring massive compute resources. This approach enables localisation and relevance to domestic contexts.
Rachel Silley pointed to open-source models as a means for developers to take existing models and tailor them to their specific needs, making them localised and relevant to their domestic context, as opposed to relying on existing larger models .
on: The compute divide between developed and developing countries is a significant challenge that requires both investment and alternative technological approaches
on: How to address the compute divide facing the Global South — through direct investment and access to large-scale compute, or through alternative low-compute and frugal AI approaches
UNGIS serves as a coordination mechanism for 50 UN agencies on digital cooperation, and member states have now mandated it to engage more actively with non-governmental stakeholders
Arg. 1Gitanjali Sah explained that UNGIS was originally created as an interagency coordination mechanism within the UN system, but member states have now directed it to broaden its engagement to include non-governmental stakeholders. This shift reflects the growing recognition of multi-stakeholderism in digital governance.
Gitanjali Sah described UNGIS as having a smooth, friendly, and productive collaboration among 50 UN agencies on digital cooperation . She noted that member states have said that UNGIS should be engaging with stakeholders, a result she attributed in part to the participation of the private sector and others in the WSIS Plus 20 process . The commitment to this was demonstrated at the Secretary General-hosted UNGIS high-level lunch, attended by 10 heads of UN agencies, at least 15 deputies, and around 50 UN agencies .
on: Multi-stakeholderism is the essential foundation of the WSIS process and must be preserved
on: Whether multi-stakeholder conversations under WSIS and the Global Digital Compact have translated into meaningful, concrete action or remain largely rhetorical
A new Multi-Stakeholder Consultation Group is being established within UNGIS, with around 200 nominations received, representing both experienced practitioners and new participants in the WSIS process
Arg. 2Gitanjali Sah announced the creation of a Multi-Stakeholder Consultation Group under UNGIS, which received an unexpectedly large number of nominations from across the stakeholder community. The group is intended to be consultative rather than interfering in day-to-day business, and may operate at international, regional, and national levels.
UNGIS launched a nomination process for the multi-stakeholder consultation group and received around 200 nominations, far exceeding expectations, from a diverse range of individuals including both long-standing WSIS participants and new faces . One model being considered is a core group at the international level with representations at regional and national levels . The group is explicitly intended to be consultative and not interfering in day-to-day business .
on: Translating WSIS commitments and high-level digital cooperation outcomes into concrete, actionable guidance remains an unresolved challenge
A digital finance task force is being created under UNGIS, bringing together experts from OECD, World Bank, UNCTAD, ITU, UNESCO, and UNDP to explore gaps in digital finance in developing countries
Arg. 3Gitanjali Sah described a new digital financial task force being established under UNGIS, drawing on expertise from a wide range of international organisations. The task force will examine gaps in digital finance in developing countries and will consult with the private sector and multilateral banks as part of its mandate.
The digital financial task force has already seen nominations of experts from organisations including OECD, World Bank, UNCTAD, ITU, UNESCO, and UNDP . Its mandate is to explore gaps in digital finance in developing countries, and the private sector, multilateral banks, and regional banks will all be consulted as part of its terms of reference .
Post-COVID, there is a global recognition that digital technologies and AI are essential to achieving development outcomes, making enabling policy environments more critical than ever
Arg. 1Yu Ping Chan argued that the COVID-19 pandemic accelerated a global realisation that digital technologies and AI are indispensable for achieving development outcomes, elevating the importance of enabling policy environments. This recognition gives new significance to the WSIS framework's emphasis on policy environments.
Yu Ping Chan noted that the acceleration in interest around digital and AI came only after COVID, where there is now a realisation that digital technologies and AI are needed to achieve development-type outcomes . She observed that when WSIS was first conceptualised 20 years ago, notions around enabling policy environments were considered more nascent and fringe, but now digital technologies and AI are ubiquitous in almost every part of society, including increasingly in developing countries .
UNDP supports over 130 governments in using digital and AI to achieve the SDGs, demonstrating the scale at which enabling environments must be built, particularly in developing countries
Arg. 2Yu Ping Chan highlighted UNDP's extensive on-the-ground presence and its work supporting governments in harnessing digital and AI for sustainable development. This scale of engagement underscores the critical need for robust enabling environments, especially in developing countries making foundational choices about digital ecosystems.
UNDP is present in 170 countries around the world and supports over 130 governments on using digital and AI to achieve the SDGs . Yu Ping Chan noted that for many developing countries, they are making critical choices around the use of digital and AI for the first time, building digital ecosystems across entire societies and communities .
The multi-stakeholder conversations initiated through WSIS must move beyond dialogue into concrete guidance and accountability mechanisms for both private sector and national governments
Arg. 3Yu Ping Chan expressed concern that despite years of multi-stakeholder conversations under WSIS and the Global Digital Compact, these have not yet translated into concrete guidance or accountability mechanisms. She called for a move from words on paper to actionable outcomes that change behaviour on the ground.
Yu Ping Chan questioned whether the conversations around the global digital compact and WSIS have landed in anything more meaningful than just conversations where the importance of multi-stakeholderism is recognised, without reaching the point of landing in action that guides what businesses and governments do . She described struggling with how to translate words on paper and conversations into concrete guidance for both private sector and national governments, with civil society as part of a continuous feedback loop and accountability mechanism .
on: Translating WSIS commitments and high-level digital cooperation outcomes into concrete, actionable guidance remains an unresolved challenge
on: Whether multi-stakeholder conversations under WSIS and the Global Digital Compact have translated into meaningful, concrete action or remain largely rhetorical
Translating WSIS commitments and global digital compact outcomes into concrete, actionable guidance for businesses, governments, and civil society remains an unresolved challenge
Arg. 4Yu Ping Chan identified the gap between high-level commitments and on-the-ground implementation as a persistent and unresolved challenge. She called for concrete outcomes and guidance that can be given to national governments and multinational corporations to actually change conditions on the ground.
Yu Ping Chan asked what the next step is after the WSIS Plus 20 outcome and global digital compact, and how to get more concrete in terms of the things that are done . She emphasised UNDP's focus on in-country implementation and the need to draw concrete things out of these processes to give to national governments and global tech operating in developing countries such that it actually changes behaviour and conditions on the ground .
on: Translating WSIS commitments and high-level digital cooperation outcomes into concrete, actionable guidance remains an unresolved challenge
UNDP has brokered public-private partnerships providing GPU access and cloud credits to African startups, demonstrating a practical model for giving the Global South access to AI compute resources
Arg. 5Yu Ping Chan provided a concrete example of how UNDP has facilitated access to AI compute resources for African startups through public-private partnerships, involving both government and private sector actors. This model is presented as a replicable approach to addressing the compute divide.
Through the AI Help for Sustainable Development initiative with the Italian government as part of the G7 presidency, UNDP brokered a deal where Seneca, the Italian supercomputer, provided around 1.5 million GPU hours for 130 African startups, while Microsoft and AWS provided approximately three million dollars' worth of cloud credits . These resources are already being deployed across 50 startups in sectors like health, agriculture, energy, and multilingual AI languages, focusing on 14 African countries outlined by Italy's Mattei Plan .
on: The compute divide between developed and developing countries is a significant challenge that requires both investment and alternative technological approaches
on: How to address the compute divide facing the Global South — through direct investment and access to large-scale compute, or through alternative low-compute and frugal AI approaches
The technical community acknowledges and appreciates the business community's engagement in the WSIS Plus 20 review, noting that without business participation the outcomes could have been very different
Arg. 1The audience speaker, representing the technical community at ICANN, placed on record the gratitude of the technical community to the business community for their engagement in the WSIS Plus 20 review process. He argued that business participation was instrumental in shaping the outcomes of this intergovernmental conversation.
Jan Skarklin, Head of Government Engagement at ICANN, stated that he wanted to place on record the gratitude of the technical community to the business community for their engagement in the WSIS Plus 20 review process and participation in placing the voice of the business community in the intergovernmental conversation . He noted that without the participation of the business community, the results might have been very different . He also encouraged the business community to participate in the preparations for and at IGF in Nairobi with full strength .
on: Multi-stakeholderism is the essential foundation of the WSIS process and must be preserved
Three key areas where the private sector can contribute to digital development financing are: transparency about local value capture, collaboration on interoperability, and supporting local businesses in developing countries for digital exports
Arg. 1Sebastian Mhatre outlined three specific areas where the private sector can make meaningful contributions to digital development financing. These areas focus on transparency, interoperability, and enabling developing countries to participate more fully in the digital economy.
Sebastian Mhatre identified three areas: first, being more transparent about what constitutes local value capture; second, working with the multi-stakeholder community around interoperability; and third, supporting local businesses within developing countries for digital exports .
on: Public-private partnerships and blended financing models are necessary to address the scale of digital development investment needs
Developing countries face structural challenges in competing in the digital economy due to their position low on the value chain, and private sector partnerships with local businesses could help shift this dynamic
Arg. 2Sebastian Mhatre argued that many developing countries are currently limited to exporting raw materials or adding small amounts of value, and digital technology offers an opportunity to move up the value chain. Private sector partnerships with local businesses could be a key mechanism for enabling this shift.
Sebastian Mhatre noted that most developing countries are currently unable to compete in the digital economy space, often only exporting raw materials or adding small amounts of value to those raw materials . He suggested that if private sector actors partner with local businesses to support them in exporting, this creates an opportunity for driving digital exports for developing countries in a new area where they previously had not had that opportunity .
There is a significant divide in compute resources between developed and developing countries, and bridging it requires massive investment that even advanced economies struggle to mobilise
Arg. 3Sebastian Mhatre acknowledged the scale of the compute divide as a fundamental challenge, noting that even advanced economies struggle to invest sufficiently in digital infrastructure. He argued that this makes it very difficult to see how the gap can be bridged through conventional investment approaches alone.
Sebastian Mhatre stated that when talking about the divide in terms of compute, one is talking about a massive divide in terms of investment, and that every country in the world is struggling with the ability to invest in their digital infrastructure, making it very difficult to see how the gap is bridged .
on: The compute divide between developed and developing countries is a significant challenge that requires both investment and alternative technological approaches
on: How to address the compute divide facing the Global South — through direct investment and access to large-scale compute, or through alternative low-compute and frugal AI approaches
Alternative approaches such as edge computing, locally run small language models, and frugal AI tailored to specific use cases can create innovation ecosystems in the Global South without requiring the same scale of compute as advanced economies
Arg. 4Sebastian Mhatre proposed that rather than solely pursuing parity with advanced economies in compute resources, the Global South could explore alternative technological approaches that are less compute-intensive. These include edge computing, small language models, and frugal AI tailored to specific local use cases.
Sebastian Mhatre suggested looking at low compute options, edge compute options, and locally run small language models as alternative approaches that can create opportunities for a different type of innovation ecosystem in the Global South . He noted that while compute will still be needed, these approaches mean developing countries are not quite as dependent on the same scale of compute as most advanced economies, and described this as a really exciting area that needs more work but could have a lot of value .
on: The compute divide between developed and developing countries is a significant challenge that requires both investment and alternative technological approaches
on: How to address the compute divide facing the Global South — through direct investment and access to large-scale compute, or through alternative low-compute and frugal AI approaches
AWS and Amazon have made significant investments in capacity building and data centre deployment in the Global South, including in India and Brazil, and are working with governments on AI and e-government initiatives
Arg. 1The AWS/Amazon representative highlighted Amazon's investments in the Global South, including data centre deployment and capacity building, as well as collaboration with governments on AI and e-government initiatives. This was offered as a response to concerns about the compute divide facing developing countries.
The AWS/Amazon representative noted significant investment in capacity building within the Global South as well as in the deployment of data centres, particularly in countries such as India and Brazil, and also working locally with governments on e-government initiatives to get governments up to speed on AI initiatives .
on: The compute divide between developed and developing countries is a significant challenge that requires both investment and alternative technological approaches
on: How to address the compute divide facing the Global South — through direct investment and access to large-scale compute, or through alternative low-compute and frugal AI approaches
The environmental costs of AI, including its carbon footprint from enormous computing power and energy consumption, must be addressed alongside innovation and profitability goals
Arg. 1Audience member Ahmed Khaled Gamal, as relayed through the moderator, raised the concern that the environmental costs of AI are being overlooked in discussions focused on innovation and investment. He called for businesses to balance innovation and profitability with the responsibility to build energy-efficient and climate-conscious AI systems.
The comment, shared via the chat and relayed by the moderator, noted that the real challenge is not only mobilising funding and investment for AI and digital transformation, but also addressing the environmental costs of these technologies, as AI requires enormous computing power and energy and its growing carbon footprint cannot be ignored . The question posed was how businesses can balance innovation and profitability with their responsibility to build AI systems that are energy efficient, climate conscious, and aligned with global sustainability goals .
The ICC Basis Workshop aims to present the role of business in contributing to the WSIS Plus 20 follow-up process and advancing broader digital cooperation efforts, including those related to the Global Digital Compact
Arg. 1Meni Anastasiadou framed the purpose of the session as highlighting how business can support the delivery of WSIS action lines and contribute to digital cooperation efforts following the adoption of the WSIS Plus 20 outcome document in December. The session is intended to demonstrate the concrete and proactive role of the private sector in translating WSIS commitments into scalable initiatives.
The goal of the conversation was described as presenting the role of business in contributing to the follow-up process since the WSIS Plus 20 outcome document was adopted in December, and how business is supporting the delivery of relevant action lines and advancing broader digital cooperation efforts, including those related to the Global Digital Compact . The session also aimed to highlight opportunities for the private sector to actively and proactively engage with the UNGIS group and UN partners to translate WSIS commitments into concrete, scalable initiatives through sustained collaboration .
on: Translating WSIS commitments and high-level digital cooperation outcomes into concrete, actionable guidance remains an unresolved challenge
The WSIS Plus 20 Outcome Document has called for the ITU to create a task force on financial mechanisms for digital development, and the private sector has a potential role to play in this process
Arg. 2Meni Anastasiadou highlighted that the WSIS Plus 20 Outcome Document specifically mandated the creation of a task force to examine financial mechanisms for digital development. She framed this as an opportunity to explore what role the private sector could play, inviting both business and government perspectives on the matter.
Anastasiadou noted that the WSIS Plus 20 Outcome Document called for the ITU to create a task force that would look into financial mechanisms for digital development, and that this leads to a discussion of what role the private sector could eventually play and what it means for the private sector . She invited both business and government perspectives on this topic .
The environmental costs of AI, including its carbon footprint from enormous computing power and energy consumption, must be addressed alongside innovation and profitability goals
Arg. 3Anastasiadou relayed a concern raised in the chat by audience member Ahmed Khaled Gamal, noting that the real challenge is not only mobilising funding and investment for AI and digital transformation, but also addressing the environmental costs of these technologies. She highlighted the question of how businesses can balance innovation and profitability with the responsibility to build energy-efficient and climate-conscious AI systems aligned with global sustainability goals.
Anastasiadou shared that a colleague in the chat noted the real challenge is not only mobilising funding and investment for AI and digital transformation, but also the need to address the environmental costs of these technologies, as AI requires enormous computing power and energy and its growing carbon footprint cannot be ignored . The question posed was how businesses can balance innovation and profitability with their responsibility to build AI systems that are energy efficient, climate conscious, and aligned with global sustainability goals .
Developing countries are concerned about being excluded from the AI production ecosystem due to significant gaps in compute resources and infrastructure
Arg. 1Audience member Jimson Olufouye raised the concern that the Global South risks being left behind in the AI production ecosystem, not merely in terms of access to the internet or general technologies, but specifically in terms of access to the compute resources required to participate in AI development. He questioned whether initiatives exist to enable developers in developing countries to access the vast computational infrastructure available in global networks.
Jimson Olufouye expressed concern about how developing countries can be inclusive in the AI production ecosystem, noting the gap in compute resources and infrastructure . He asked whether Meta was planning or coming up with an initiative that would enable developing countries and their developers to enjoy the vast infrastructural computing resources available in global networks .
on: The compute divide between developed and developing countries is a significant challenge that requires both investment and alternative technological approaches
Session Knowledge Graph
Speakers · Topics · Arguments · Relationships
All speakers converged on the view that the WSIS process must remain genuinely multi-stakeholder. Helen Harris noted that none of Amazon's concrete contributions happen without the multi-stakeholder table , and that the business community worked to preserve the consensus-based approach during the WSIS Plus 20 review . Gitanjali Sah confirmed that WSIS was conceived as inherently multi-stakeholder from its inception , and that member states have now directed UNGIS to broaden engagement beyond UN agencies . Rachel Silley argued that inclusive decision-making leads to robust, practical, and globally accepted solutions . Yu Ping Chan called for these conversations to move into the mainstream . Jan Skarklin from ICANN placed on record the technical community's gratitude to business for their engagement, noting outcomes could have been very different without them .
Business community's participation in the WSIS Plus 20 review process was instrumental in preserving the consensus-based, multi-stakeholder approach to internet governance
UNGIS serves as a coordination mechanism for 50 UN agencies on digital cooperation, and member states have now mandated it to engage more actively with non-governmental stakeholders
The private sector brings unique real-time expertise in innovation, infrastructure, and technical standards, making its inclusion in financing discussions essential for robust and practical solutions
The multi-stakeholder conversations initiated through WSIS must move beyond dialogue into concrete guidance and accountability mechanisms for both private sector and national governments
The technical community acknowledges and appreciates the business community's engagement in the WSIS Plus 20 review, noting that without business participation the outcomes could have been very different
Business representatives from Amazon, Meta, and AWS all agreed that private sector investment in infrastructure and capacity building is a concrete mechanism for advancing WSIS goals. Helen Harris cited Amazon's Project Kuiper satellite constellation of 3,236 planned satellites , training of over 31 million people , and delivery of 26 million relief items . Rachel Silley highlighted Meta's subsea cable infrastructure commitment through the ITU's Partners to Connect Coalition and its open AI tools applied across education and agriculture . The AWS representative noted significant investment in data centres in India and Brazil and e-government initiatives .
Private sector contributions to WSIS action lines through infrastructure, capacity building, and disaster relief - Amazon has deployed satellite broadband (Project Kuiper), trained 31 million people in cloud and AI skills, and delivered 26 million relief items across 200+ disasters
Meta's investments in subsea cable infrastructure, workforce skills academies, and open-source AI tools directly advance WSIS goals around connectivity, capacity building, and inclusive digital transformation
AWS and Amazon have made significant investments in capacity building and data centre deployment in the Global South, including in India and Brazil, and are working with governments on AI and e-government initiatives
Multiple speakers agreed that neither public nor private financing alone can meet the scale of digital development needs, and that blended public-private models are essential. Rachel Silley explicitly stated that private capital deployment in subsea cables is the kind of investment that public financing alone cannot match at scale, and cited this as the blended cooperation the WSIS Plus 20 outcome document calls for . Yu Ping Chan provided a concrete example through UNDP's AI Help for Sustainable Development initiative, where Seneca provided 1.5 million GPU hours and Microsoft and AWS provided approximately three million dollars in cloud credits for African startups . Sebastian Mhatre outlined three areas for private sector contribution including transparency, interoperability, and supporting digital exports . Helen Harris called for UNGIS to serve as a single coordination point to make business-UN collaboration more efficient .
UNGIS should function as a single entry point for business engagement with the UN system, reducing the burden of navigating multiple separate agencies
Private capital deployment in backbone infrastructure, such as subsea cables, complements public financing and represents the kind of blended public-private cooperation explicitly called for in the WSIS Plus 20 outcome document
UNDP has brokered public-private partnerships providing GPU access and cloud credits to African startups, demonstrating a practical model for giving the Global South access to AI compute resources
Three key areas where the private sector can contribute to digital development financing are: transparency about local value capture, collaboration on interoperability, and supporting local businesses in developing countries for digital exports
Audience member Jimson Olufouye raised the concern about the compute divide facing developing countries , which prompted broad agreement among panellists that this is a real and serious challenge. Sebastian Mhatre acknowledged the massive investment divide and the difficulty of bridging it through conventional means , while proposing alternative approaches including edge computing and small language models . Yu Ping Chan agreed, pointing to frugal AI and smart AI as pragmatic approaches , and provided the concrete example of UNDP brokering GPU access and cloud credits for African startups . Rachel Silley pointed to open-source models as enabling localisation without requiring massive compute . The AWS representative highlighted Amazon's data centre investments in India and Brazil .
There is a significant divide in compute resources between developed and developing countries, and bridging it requires massive investment that even advanced economies struggle to mobilise
Alternative approaches such as edge computing, locally run small language models, and frugal AI tailored to specific use cases can create innovation ecosystems in the Global South without requiring the same scale of compute as advanced economies
UNDP has brokered public-private partnerships providing GPU access and cloud credits to African startups, demonstrating a practical model for giving the Global South access to AI compute resources
Open-source AI models allow developers in developing countries to tailor existing models to their local context, reducing dependency on large-scale compute infrastructure
AWS and Amazon have made significant investments in capacity building and data centre deployment in the Global South, including in India and Brazil, and are working with governments on AI and e-government initiatives
Developing countries are concerned about being excluded from the AI production ecosystem due to significant gaps in compute resources and infrastructure
Yu Ping Chan most explicitly articulated the challenge of moving from words to action, questioning whether WSIS and Global Digital Compact conversations have landed in anything more meaningful than just conversations , and calling for concrete guidance for private sector, national governments, and civil society . She asked what the next step is after WSIS Plus 20 and how to get more concrete , emphasising UNDP's focus on in-country implementation . Gitanjali Sah responded by describing the new Multi-Stakeholder Consultation Group with around 200 nominations and the digital finance task force as concrete mechanisms. Meni Anastasiadou framed the entire session around this challenge, noting the goal of translating WSIS commitments into concrete, scalable initiatives .
The multi-stakeholder conversations initiated through WSIS must move beyond dialogue into concrete guidance and accountability mechanisms for both private sector and national governments
Translating WSIS commitments and global digital compact outcomes into concrete, actionable guidance for businesses, governments, and civil society remains an unresolved challenge
A new Multi-Stakeholder Consultation Group is being established within UNGIS, with around 200 nominations received, representing both experienced practitioners and new participants in the WSIS process
The ICC Basis Workshop aims to present the role of business in contributing to the WSIS Plus 20 follow-up process and advancing broader digital cooperation efforts, including those related to the Global Digital Compact
Both Helen Harris and Rachel Silley presented their respective companies' contributions as concrete, measurable implementations of WSIS action lines, demonstrating that private sector engagement translates into tangible outcomes across connectivity, skills, and inclusive digital transformation. Harris cited Amazon's satellite broadband constellation , training of 31 million people , and disaster relief operations . Silley cited Meta's subsea cable infrastructure , Workforce Academy , and open AI tools applied in agriculture and healthcare . Both framed these as examples of the blended public-private cooperation called for in the WSIS Plus 20 outcome document. Both Sebastian Mhatre and Yu Ping Chan agreed that addressing the AI compute divide in the Global South requires both practical partnership models and alternative technological approaches that do not simply replicate the resource-intensive models of advanced economies. Mhatre proposed low compute options, edge computing, and small language models as creating a different type of innovation ecosystem . Chan agreed, pointing to frugal AI and smart AI tailored to specific use cases , while also providing the concrete UNDP example of brokering GPU access and cloud credits for African startups . Both emphasised pragmatism over parity. Both Gitanjali Sah and Helen Harris shared the view that UNGIS should serve as a central coordination point that bridges the UN system and the private sector. Harris called for UNGIS to be the single door business knocks on instead of a dozen separate ones , and noted it can encourage collaboration with industry so business input feeds directly into action line work . Sah confirmed that member states have now directed UNGIS to engage with stakeholders beyond the UN system , and described the high-level commitment demonstrated at the Secretary General-hosted UNGIS lunch . Both saw this as a structural improvement to multi-stakeholder engagement. Both Yu Ping Chan and Rachel Silley agreed that the accelerating pace of digital transformation and AI adoption makes inclusive, multi-stakeholder engagement in policy and financing discussions more urgent than ever. Chan noted that post-COVID, digital and AI are now ubiquitous and the need for enabling environments has grown significantly , and that the multi-stakeholder conversation must move into the mainstream . Silley argued that the technology sector's unique real-time expertise makes its inclusion essential , and that inclusive decision-making leads to robust, practical, and globally accepted solutions . Both framed this as a matter of urgency given the accelerating pace of change . Harris, Silley, and Chan all shared the view that the private sector's role in WSIS and digital governance is not merely consultative but substantive and essential. Harris emphasised that business worked to preserve the consensus-based approach and that governance (Action Line C1) underpins all other contributions . Silley argued that engagement is not a courtesy but because inclusive decision-making leads to better outcomes . Chan called for this engagement to move beyond conversations into concrete guidance and accountability , reflecting a shared frustration with the gap between rhetoric and action.
It was somewhat unexpected that a UK government representative, a UN development agency, and a major technology company all converged on the view that the Global South should not simply seek to replicate the compute-intensive AI models of advanced economies, but should instead pursue frugal, small, and locally tailored AI approaches. This represents a notable departure from the typical narrative of simply closing the digital divide by replicating existing infrastructure. Mhatre proposed low compute options, edge computing, and small language models . Chan agreed, pointing to frugal AI and smart AI as pragmatic approaches . Silley pointed to open-source models as enabling localisation . This consensus across government, UN, and private sector on a nuanced, context-specific approach to AI development was not anticipated given the different institutional perspectives represented.
It was somewhat unexpected that Yu Ping Chan from UNDP, a core UN agency deeply embedded in the WSIS process, openly acknowledged that the conversations around WSIS and the Global Digital Compact may not have landed in anything more meaningful than just conversations , and expressed personal struggle with translating words on paper into concrete guidance . This candid self-criticism from within the UN system, at a session hosted by the ITU and ICC to celebrate WSIS achievements, was notable. Gitanjali Sah's response by pointing to the new Multi-Stakeholder Consultation Group and digital finance task force as concrete mechanisms implicitly acknowledged the validity of Chan's concern, representing an unexpected moment of institutional self-reflection and consensus on the need for greater concreteness.
The environmental sustainability of AI emerged as an unexpected point of consensus when audience member Ahmed Khaled Gamal raised the issue via the chat , and moderator Meni Anastasiadou not only relayed it but implicitly endorsed it by noting it had been addressed in earlier examples and inviting further responses . In a session dominated by enthusiasm for AI investment and deployment, the fact that this concern was raised and acknowledged without pushback from any panellist represents an unexpected area of consensus that environmental costs cannot be ignored alongside innovation and profitability goals .
The discussion revealed a high degree of consensus across all speaker categories - UN agencies, private sector, government, and technical community - on several core themes: the indispensability of multi-stakeholderism in digital governance ; the concrete and measurable nature of private sector contributions to WSIS action lines ; the necessity of blended public-private financing models ; the urgency of addressing the AI compute divide in the Global South through both investment and alternative approaches ; and the persistent challenge of translating high-level commitments into concrete, actionable outcomes . There was also unexpected consensus on the need for frugal and context-specific AI approaches for developing countries , and on the importance of addressing the environmental costs of AI .
Yu Ping Chan expressed significant scepticism about whether WSIS and Global Digital Compact processes have moved beyond conversation, questioning whether they have 'landed in anything any more meaningful than just conversations' and whether they have reached 'the point now of actually landing it in action' . She described struggling with 'how to translate words on paper, conversations like this, into then what is concrete guidance' and asked 'what is the next step after WSISP20 outcome, global digital compact and so forth' . By contrast, Helen Harris presented multiple concrete deliverables across WSIS action lines , Rachel Silley cited specific infrastructure and skills investments , and Gitanjali Sah outlined new institutional mechanisms including the Multi-Stakeholder Consultation Group and digital finance task force , all of which implicitly pushed back against the characterisation that progress has been merely rhetorical.
The multi-stakeholder conversations initiated through WSIS must move beyond dialogue into concrete guidance and accountability mechanisms for both private sector and national governments
Private sector contributions to WSIS action lines through infrastructure, capacity building, and disaster relief - Amazon has deployed satellite broadband (Project Kuiper), trained 31 million people in cloud and AI skills, and delivered 26 million relief items across 200+ disasters
Meta's investments in subsea cable infrastructure, workforce skills academies, and open-source AI tools directly advance WSIS goals around connectivity, capacity building, and inclusive digital transformation
UNGIS serves as a coordination mechanism for 50 UN agencies on digital cooperation, and member states have now mandated it to engage more actively with non-governmental stakeholders
Sebastian Mhatre acknowledged the compute divide as a 'massive divide in terms of investment' but argued that conventional approaches to bridging it are unrealistic, instead advocating for 'low compute options, edge compute options, locally run small language models' and 'frugal AI' tailored to specific use cases . Yu Ping Chan, while endorsing the frugal AI direction , also presented a concrete model of brokering direct GPU access and cloud credits for African startups through public-private partnerships , suggesting a more direct approach to compute access is also viable. Rachel Silley focused on open-source models as a means of reducing dependency on large-scale compute , while the AWS representative emphasised direct infrastructure investment in data centres in countries like India and Brazil . These responses reflect differing emphases on whether the solution lies in working around the compute gap or in directly addressing it.
There is a significant divide in compute resources between developed and developing countries, and bridging it requires massive investment that even advanced economies struggle to mobilise
Alternative approaches such as edge computing, locally run small language models, and frugal AI tailored to specific use cases can create innovation ecosystems in the Global South without requiring the same scale of compute as advanced economies
UNDP has brokered public-private partnerships providing GPU access and cloud credits to African startups, demonstrating a practical model for giving the Global South access to AI compute resources
Open-source AI models allow developers in developing countries to tailor existing models to their local context, reducing dependency on large-scale compute infrastructure
AWS and Amazon have made significant investments in capacity building and data centre deployment in the Global South, including in India and Brazil, and are working with governments on AI and e-government initiatives
It was notably unexpected for Yu Ping Chan, representing UNDP as Vice Chair of UNGIS, to openly question whether the WSIS and Global Digital Compact processes have produced anything more than 'talking point bingo where you have a couple of meetings and that pretty much just becomes all that is' . She described herself as 'struggling with that part of how do you translate words on paper, conversations like this, into then what is concrete guidance' and questioned whether the processes have 'actually landed it in anything any more meaningful than just conversations' . This candid scepticism from within the UN system, at a workshop explicitly designed to celebrate and advance WSIS outcomes, stood in sharp contrast to the broadly affirmative tone of the other panellists and was unexpected given the institutional context. It implicitly challenged the framing of the entire session, which was premised on the value of the WSIS Plus 20 outcomes and their follow-up .
Throughout the discussion, AI was consistently framed by the main panellists as an accelerant of economic growth and development , with investments in AI infrastructure and accessibility presented as unambiguously positive contributions to WSIS goals. The environmental costs of AI were not raised by any of the main panellists. It was therefore unexpected when an audience member, relayed by the moderator, challenged this framing by noting that 'the real challenge is not only mobilising funding and investment for AI and digital transformation' but also addressing 'the environmental costs of these technologies' and AI's 'growing carbon footprint' . The moderator acknowledged the point but noted it had 'been addressed in the examples that you've shared previously' , a response that itself was somewhat dismissive given that none of the panellists had substantively addressed the environmental dimension of their AI investments. This introduced an unexpected tension between the pro-AI investment narrative and sustainability concerns.
The discussion was broadly collaborative and consensus-oriented, with most speakers aligned on the importance of multi-stakeholderism, private sector engagement, and the WSIS framework. The most substantive disagreement emerged between Yu Ping Chan's candid scepticism about whether WSIS processes have produced concrete outcomes and the other panellists' presentation of tangible deliverables . A secondary disagreement concerned the best approach to addressing the compute divide in the Global South, with speakers variously emphasising direct infrastructure investment , frugal and low-compute alternatives , brokered public-private partnerships for compute access , and open-source model adaptation . An unexpected challenge to the dominant pro-AI narrative came from an audience member raising environmental costs , which was not substantively addressed by the panel.
All speakers agreed that multi-stakeholderism is essential and that the private sector must be actively engaged in WSIS implementation and digital governance processes. Helen Harris framed the multi-stakeholder table as the foundation without which none of the concrete contributions happen , Gitanjali Sah confirmed that WSIS was designed to be multi-stakeholder from its inception and that member states have now mandated UNGIS to engage with stakeholders , Rachel Silley argued that inclusive decision-making leads to 'robust, practical, and globally accepted solutions' , and Sebastian Mhatre's three-point framework implicitly assumed multi-stakeholder collaboration . However, Yu Ping Chan diverged on whether this shared commitment to multi-stakeholderism has yet produced concrete, actionable outcomes, questioning whether the conversations have moved beyond recognition of the principle into actual guidance and accountability . Thus all agreed on the goal of meaningful multi-stakeholder engagement but disagreed on whether it has been achieved.
Business community's participation in the WSIS Plus 20 review process was instrumental in preserving the consensus-based, multi-stakeholder approach to internet governance UNGIS serves as a coordination mechanism for 50 UN agencies on digital cooperation, and member states have now mandated it to engage more actively with non-governmental stakeholders The multi-stakeholder conversations initiated through WSIS must move beyond dialogue into concrete guidance and accountability mechanisms for both private sector and national governments The private sector brings unique real-time expertise in innovation, infrastructure, and technical standards, making its inclusion in financing discussions essential for robust and practical solutions Three key areas where the private sector can contribute to digital development financing are: transparency about local value capture, collaboration on interoperability, and supporting local businesses in developing countries for digital exports
All speakers agreed that the compute divide is a real and serious problem that risks excluding the Global South from the AI production ecosystem, as articulated by Jimson Olufouye's question and acknowledged by Sebastian Mhatre , Yu Ping Chan , Rachel Silley , and the AWS representative . However, they disagreed on the primary solution: Sebastian Mhatre emphasised frugal and low-compute alternatives , Yu Ping Chan highlighted direct brokering of GPU and cloud credit access , Rachel Silley pointed to open-source model adaptation , and the AWS representative focused on infrastructure investment . The shared goal of AI inclusion for the Global South was clear, but the pathways proposed differed meaningfully.
There is a significant divide in compute resources between developed and developing countries, and bridging it requires massive investment that even advanced economies struggle to mobilise UNDP has brokered public-private partnerships providing GPU access and cloud credits to African startups, demonstrating a practical model for giving the Global South access to AI compute resources Open-source AI models allow developers in developing countries to tailor existing models to their local context, reducing dependency on large-scale compute infrastructure AWS and Amazon have made significant investments in capacity building and data centre deployment in the Global South, including in India and Brazil, and are working with governments on AI and e-government initiatives Developing countries are concerned about being excluded from the AI production ecosystem due to significant gaps in compute resources and infrastructure
All three speakers agreed that enabling policy environments and blended public-private investment are essential for digital transformation, particularly in developing countries. Yu Ping Chan noted that post-COVID, the need for government or public sector intervention to harness digital technologies 'becomes even more important' and that this is 'even more critical in developing countries' . Helen Harris and Rachel Silley both presented concrete examples of private sector investment filling gaps that public financing cannot address at scale . However, Yu Ping Chan diverged by questioning whether these investments and the policy conversations around them are being translated into concrete guidance that actually changes 'behaviour and conditions on the ground' , whereas Harris and Silley presented their contributions as already delivering tangible outcomes.
Post-COVID, there is a global recognition that digital technologies and AI are essential to achieving development outcomes, making enabling policy environments more critical than ever Private sector contributions to WSIS action lines through infrastructure, capacity building, and disaster relief - Amazon has deployed satellite broadband (Project Kuiper), trained 31 million people in cloud and AI skills, and delivered 26 million relief items across 200+ disasters Private capital deployment in backbone infrastructure, such as subsea cables, complements public financing and represents the kind of blended public-private cooperation explicitly called for in the WSIS Plus 20 outcome document
- The private sector has made substantial, concrete contributions to WSIS action lines, including Amazon's deployment of satellite broadband (Project Kuiper), training of 31 million people in cloud and AI skills, and delivery of 26 million relief items across 200+ disasters, as well as Meta's investments in subsea cable infrastructure, workforce academies, and open-source AI tools.
- The business community's participation in the WSIS Plus 20 review process was instrumental in preserving the consensus-based, multi-stakeholder approach to internet governance, with the technical community formally acknowledging that outcomes could have been very different without business engagement.
- UNGIS, as the UN's interagency coordination mechanism for 50 agencies on digital cooperation, has been mandated by member states to engage more actively with non-governmental stakeholders, and should function as a single entry point for business engagement with the UN system.
- A new Multi-Stakeholder Consultation Group (MCG) is being established within UNGIS, with approximately 200 nominations received from experienced practitioners and new participants alike, reflecting strong and broad interest in the WSIS process.
- A digital finance task force is being created under UNGIS, drawing on expertise from OECD, World Bank, UNCTAD, ITU, UNESCO, and UNDP, with a mandate to explore gaps in digital finance in developing countries and to consult the private sector as part of its terms of reference.
- Post-COVID, there is a global recognition that digital technologies and AI are essential to achieving development outcomes, making enabling policy environments more critical than ever, particularly in developing countries where foundational digital ecosystems are still being built.
- UNDP supports over 130 governments in using digital and AI to achieve the SDGs, and has brokered public-private partnerships providing GPU access and cloud credits to African startups, demonstrating a practical model for giving the Global South access to AI compute resources.
- Translating WSIS commitments and Global Digital Compact outcomes into concrete, actionable guidance and accountability mechanisms for businesses, governments, and civil society remains a central and as yet unresolved challenge.
- Three key areas where the private sector can contribute to digital development financing are: transparency about local value capture, collaboration on interoperability, and supporting local businesses in developing countries for digital exports.
- Alternative approaches such as edge computing, locally run small language models, frugal AI, and open-source models tailored to specific use cases can create innovation ecosystems in the Global South without requiring the same scale of compute infrastructure as advanced economies.
- The environmental costs of AI, including its carbon footprint from enormous computing power and energy consumption, must be addressed alongside innovation and profitability goals.
“None of it happens without the multi-stakeholder table we're sitting at right now. Which brings me to UNGIS. As the UN's own interagency mechanism for coordinating WSIS and Information Society work, UNGIS is uniquely placed to be the single door business knocks on instead of a dozen separate ones.”
“I'm struggling with that part of how do you translate words on paper, conversations like this, into then what is concrete guidance for both private sector as well as national governments... Besides having the multi-stakeholder advisory group that Gitanjali was talking about, besides conversations, what can we come to that's concrete in terms of outcomes or these guidance and expectations for what we think private sector as well as national governments and then also particularly with the role for civil society can play?”
“What you've pointed out is a really important problem but it's not one that has any easy solution because essentially when you talk about the divide in terms of compute you're talking about a massive divide in terms of investment... one of the things I'm quite interested in is some of the alternative approaches... low compute options, edge compute options, locally run small language models. All these things can create opportunities for a different type of innovation ecosystem.”
“UNDP has brokered a number of public-private partnerships to try and support the access to compute and cloud credits... through the AI help for sustainable development... Seneca which is the Italian supercomputer provided GPU access around 1.5 million GPU hours announced for 130 African startups and then Microsoft and AWS actually provided cloud credits around three million dollars worth... that's a model I think for some kind of a sustainable development partnership in this specific area.”
“Without participation of business community, the results might have been very different and that's why I think you deserve a lot of appreciation. That said, the train is going and we need all of us to be sitting on that train... I just wanted to use this opportunity to encourage business community to participate in the preparations and IGF itself with the full strength.”
How can the multi-stakeholder consultation group (MCG) being established by UNGIS be structured most effectively, given the unexpectedly large number of nominations (~200) received?
The volume and diversity of nominations exceeded expectations, raising important questions about governance structure, representation, and how to balance experienced participants with new voices. The model being considered — a core international group with regional and national representations — needs further deliberation and design.
How can the conversations and commitments from WSIS Plus 20 and the Global Digital Compact be translated into concrete, actionable guidance for private sector, national governments, and civil society, rather than remaining at the level of high-level declarations?
Yu Ping Chan explicitly raised concern that despite years of multi-stakeholder dialogue, the outcomes have not yet produced concrete behavioural change or clear expectations for actors on the ground. This is a critical gap between policy aspiration and implementation reality.
What mechanisms can be established to create a continuous feedback loop and accountability framework involving private sector, national governments, and civil society in implementing WSIS commitments?
Beyond having advisory groups and conversations, there is an unresolved question of how to hold both private sector and national governments accountable for their commitments, and how civil society can play a meaningful role in that accountability process.
How can developing countries, particularly in the Global South, gain meaningful access to compute resources and infrastructure needed to participate in the AI production ecosystem, not just as consumers but as developers and innovators?
This question highlights a fundamental and growing divide — not just in internet connectivity but in the computational infrastructure required for AI development. Without addressing this gap, developing countries risk being permanently excluded from the AI value chain.
What role can low-compute, edge-compute, and small language model approaches play in enabling an alternative AI innovation ecosystem in the Global South that is less dependent on massive infrastructure investment?
Both speakers identified this as an exciting but underdeveloped area. Given the enormous investment gap in compute infrastructure, exploring pragmatic, frugal AI approaches tailored to specific development use cases in the Global South warrants significant further research and piloting.
How can public-private partnership models — such as the UNDP AI Help for Sustainable Development initiative providing GPU hours and cloud credits to African startups — be scaled and replicated to provide broader access to compute resources for developing countries?
The UNDP model involving Italian supercomputing resources and cloud credits from Microsoft and AWS for African startups represents a promising but limited proof of concept. Further research is needed on how to scale such models sustainably and extend them to more countries and sectors.
How can open-source AI models be leveraged by developers in the Global South to build locally relevant, contextualised applications without relying on large proprietary models that require significant compute resources?
Open-source models offer a potential pathway for developers in resource-constrained environments to innovate, but the practical barriers, support structures, and governance frameworks needed to make this viable at scale in developing countries require further exploration.
How can businesses balance innovation and profitability with the responsibility to build AI systems that are energy-efficient, climate-conscious, and aligned with global sustainability goals, given AI's growing carbon footprint?
As AI adoption accelerates globally, the environmental costs — particularly energy consumption and carbon emissions from data centres — represent a significant and underaddressed challenge. This question calls for further research into sustainable AI infrastructure design and corporate accountability frameworks.
How can UNGIS serve as a more effective single point of engagement for the private sector across all UN agencies, rather than requiring businesses to navigate multiple separate agency relationships?
Helen Harris raised the idea of UNGIS acting as a 'single door' for business engagement with the UN system. The practical mechanisms, governance arrangements, and processes needed to make this a reality — rather than an aspiration — require further design and research.
How can the digital financial task force being established under UNGIS effectively identify and address gaps in digital finance for developing countries, and how will private sector and multilateral financial institutions be meaningfully consulted in that process?
The newly formed digital finance task force, involving bodies such as the OECD, World Bank, UNCTAD, ITU, UNESCO, and UNDP, is at an early stage. The scope of its mandate, its methodology for identifying gaps, and the modalities for private sector and regional bank engagement all require further definition.
How can developing countries increase local value capture from digital technologies and build competitive digital export capabilities, particularly given their current position at the lower end of the value chain?
Sebastian Mhatre highlighted that most developing countries currently export raw materials or low-value-added products, and that digital technology offers an opportunity to move up the value chain. However, the conditions, partnerships, and policy frameworks needed to realise this potential require further investigation.
How can greater transparency be achieved regarding local value capture by multinational technology companies operating in developing countries, and what standards or reporting frameworks could support this?
Sebastian Mhatre identified transparency about local value capture as a key area where the private sector could improve. The development of appropriate metrics, disclosure standards, and accountability mechanisms for this purpose represents an important area for further research and policy development.
How can interoperability between digital systems and platforms be advanced through multi-stakeholder collaboration, particularly to benefit developing countries seeking to integrate into the global digital economy?
Interoperability was identified as one of three key areas where the private sector can play a constructive role, but the specific frameworks, standards, and collaborative processes needed to achieve meaningful interoperability — especially for developing country contexts — require further work.
How can the business community sustain and strengthen its engagement in upcoming processes such as the Internet Governance Forum (IGF) in Nairobi, building on the momentum from the WSIS Plus 20 review?
Jan Skarklin explicitly encouraged the business community to participate with full strength in IGF preparations and the event itself, implying that sustaining this level of engagement is not guaranteed and requires deliberate effort and planning.
