The session brought together leaders from across the United Nations system to explore how digital transformation and AI can advance sustainable development, inclusion, and resilience, and how the UN can strengthen system-wide cooperation in pursuit of the renewed WSIS mandate to 2035 .
ITU Director General Bogdan-Martin opened by noting that with 2.2 billion people still offline, the work of the UN system remains far from complete . UPU Director General Metoki highlighted that post offices, of which around 100,000 still lack meaningful internet connectivity, serve as inclusive access points for digital services in rural areas, particularly for women and micro, small, and medium-sized enterprises . WHO Director General Dr. Tedros emphasised that the digital divide is fundamentally a health divide, and that AI must be safe, ethical, and equitable, with WHO supporting over 80 countries through its Global Digital Health Certification Network . WMO Secretary General Saulo stressed that trustworthy AI depends on open data and international cooperation, pointing to WMO's unified data policy and its WIS2 open data system, which now connects over 90 countries .
In the second panel, senior UN officials shared concrete examples of implementation. UNESCO reported training over 35,000 civil servants through its AI and Digital Transformation programme , while UNITAR described reaching nearly 600,000 learners across 105 countries annually . UNICEF presented its 'Ahead of the Storm' initiative to predict child-specific impacts of natural disasters and better enable proactive humanitarian response that uses AI to protect the 1.1 billion children exposed to climate hazards . UNICRI reported that its 'AI for Safer Children' project had supported the rescue of over 100 children and the arrest of over 250 offenders .
Across both panels, participants consistently identified trust, data governance, capacity building, and equitable access as foundational prerequisites for responsible digital transformation . The discussion concluded with a shared conviction that the WSIS process exemplifies genuine UN unity in action, with the system's collective efforts representing a model for delivering measurable impact for people worldwide .
Overall Purpose
- The discussion brings together leaders from across the United Nations system at the WSIS Forum to showcase how UN agencies are collectively advancing inclusive, trusted, and sustainable digital transformation. The two-panel session moves from high-level strategic vision to concrete implementation examples, demonstrating how the UN family is working in coordination to leverage digital technologies and AI for sustainable development, while ensuring no one is left behind.
- --
Major Discussion Points
- The UN system's collective responsibility to advance the WSIS mandate through multi-stakeholder cooperation. The panel emphasised that the past 20 years of WSIS have proven that multi-stakeholder cooperation works, and that the renewed mandate to 2035 requires the UN family to act as one. UNGIS (the UN Group on the Information Society) was highlighted as the coordinating mechanism bringing together 31 UN entities and 17 observers to deliver on shared digital goals. The WSIS Plus 20 review also assigned additional responsibilities, including work on financing for digital development. - Bridging the digital divide - including connectivity, capacity, and equity gaps - as a prerequisite for sustainable development. Multiple speakers stressed that with 2.2 billion people still offline , access alone is insufficient. The 'capacity divide' was identified as a growing challenge beyond connectivity, requiring skills, institutions, leadership, and policy capabilities. UPU highlighted that approximately 100,000 post offices lack meaningful internet connection, yet post offices serve as inclusive access points in rural areas. UNCTAD noted that in much of the developing world, digital devices are used primarily for entertainment rather than economic activity, underscoring the need to make connectivity economically meaningful. - Trustworthy, ethical, and human-centred AI as a foundation for digital transformation. Speakers consistently argued that AI must be safe, equitable, and governed responsibly. Dr. Tedros stated plainly that 'the digital divide is a health divide' , and that without closing gaps in connectivity, capacity, governance, and financing, digital transformation risks widening inequalities. WMO's Secretary General reinforced that trustworthy AI begins not with algorithms but with open data, standards, and open-source software. UNRISD emphasised that social justice cannot be an afterthought in AI governance, and that the measure of AI's success must be how it contributes to society, not merely what it can do.
- Concrete UN-led digital and AI initiatives delivering measurable impact on the ground. The second panel showcased specific implementations across the system: UNICRI's AI for Safe Children project trained over 3,500 law enforcement officers from 60 countries and contributed to saving over 100 children and arresting over 250 perpetrators ; the UN Pension Fund deployed a digital identity Proof of Life solution for 80,000 retirees across 192 countries, later expanded into a UN-wide digital identity solution joined by 16 entities ; UNESCO trained more than 35,000 civil servants through its AI and Digital Transformation in Government programme ; and UNICEF developed an AI-driven climate preparedness tool addressing the 1.1 billion children exposed to multiple climate hazards. - Strengthening UN system-wide cooperation and reform (UN80) through shared digital infrastructure and interoperability. Several speakers pointed to the importance of avoiding duplication and building shared platforms. UNICC outlined a three-pillar AI framework - an AI factory, AI gateway, and AI studio - to move solutions safely from experimentation to deployment across the system. UN Ops described multi-tenant architecture systems that allow multiple agencies to share procurement, recruitment, and grant management tools rather than building 40 separate solutions. UNDP noted collaboration with ITU on an "expertise on demand" digital platform as part of the UN80 reform agenda. ---
Overall Tone
- The overall tone of the discussion is collaborative, optimistic, and action-oriented, with an undercurrent of urgency. From the outset the discussion took place in an energetic atmosphere, acknowledging the breadth of UN participation and framing the session as a demonstration of collective strength. Throughout the first panel, the tone is largely visionary and principled - leaders articulate the 'why' of digital cooperation, frequently invoking themes of trust, solidarity, and inclusion. There are moments of candour, particularly when speakers acknowledge persistent inequalities and the risk that digital transformation could widen rather than narrow divides.
- In the second panel, the tone shifts to become more practical and evidence-driven, with speakers grounding their contributions in specific projects, statistics, and measurable outcomes. This creates a sense of momentum and credibility, reinforcing the message that the UN is not merely discussing principles but actively delivering on them.
- The closing remarks maintain a unifying and forward-looking tone, with the session concluding on a note of collective pride and shared purpose. The discussion remains consistently constructive throughout.
UN Leaders Talks at the WSIS Forum: Leadership for Inclusive, Trusted, and Sustainable Digital Transformation
#
Opening and Context
The UN Leaders Talks, held during the World Summit on the Information Society (WSIS) Forum, brought together heads and senior representatives of more than a dozen United Nations agencies across two consecutive panels to demonstrate how the UN family is collectively advancing inclusive, trusted, and sustainable digital transformation. ITU Director General Doreen Bogdan-Martin opened the session by noting the breadth of UN participation throughout the week, from the Global Dialogue and AI for Good to the UN exhibition in the AI Hall and the presentation of the UN AI Hub . She framed the discussion around a fundamental collective responsibility: it is the UN family that leads as facilitators of the WSIS action lines and forms the UN Group on the Information Society (UNGIS), and it is their shared duty to carry the WSIS vision and its renewed mandate forward to 2035 .
Bogdan-Martin underscored that the past twenty years had proven multi-stakeholder cooperation works, with governments, civil society, academia, the private sector, international organisations, and the UN system together building a time-tested platform for driving progress towards a better digital future for all . She identified this moment as more crucial than ever, posing a series of strategic questions: how to leverage digital technologies to deliver on mandates, how to serve member states better, and how to ensure technology serves as an enabler for the UN reform process . Against this backdrop, she noted that with 2.2 billion people still offline, the work of the UN system remains far from complete .
---
#
Panel One: Strategic Vision from UN Leaders
##
Universal Postal Union - Post Offices as Digital Access Points
UPU Director General Masahiko Metoki opened the substantive discussion by challenging the assumption that the postal sector is old-fashioned and disconnected from the digital age . He pointed to the WSIS Plus 20 document, approved by the UN General Assembly, which explicitly recognises post offices as inclusive access points for delivering digital services in rural areas . With between 50,000 and 60,000 post offices worldwide, the majority located in rural areas, the postal sector occupies a unique position in bridging the digital divide . Post offices can provide digital services cost-effectively to people who may not easily access the internet, smartphones, or digital skills, and many postal operators now offer e-commerce, e-governance, and digital financial services in person at post office counters . Initiatives such as Post for Women empower women-led business enterprises in rural areas to build digital skills and sell products online .
Metoki acknowledged, however, that a significant gap persists: approximately 100,000 post offices worldwide still lack meaningful internet connectivity . He committed UPU to continuing efforts to connect communities through the post office, both through its own projects and in partnership with other UN organisations, to bridge the digital divide and ensure no one is left behind .
##
World Health Organisation - The Digital Divide as a Health Divide
WHO Director General Dr. Tedros Adhanom Ghebreyesus affirmed that the renewal of the WSIS mandate to 2035 sends an important message: digital cooperation remains essential to achieving sustainable development . For WHO, he argued, the question is no longer whether technology will transform health - it already is - but rather how to ensure that transformation is trusted, equitable, and responsive to people's needs . Digital transformation, he emphasised, is not about technology alone but about helping countries build stronger health systems and enabling people everywhere to access better health services .
Dr. Tedros described WHO's work as co-facilitator of the WSIS action line on e-health, supporting countries to build the foundations of digital health: interoperable systems, trusted data governance, digital public infrastructure, and the workforce capabilities needed to use these tools effectively . Through the Global Initiative on Digital Health and the Global Digital Health Partnership, WHO is bringing governments, development partners, civil society, and industry together around country-led priorities, reducing fragmentation . He highlighted that more than 80 countries, representing over 2 billion people, participate in WHO's Global Digital Health Certification Network, enabling trusted health credentials across borders, and that more than 100 countries use WHO's epidemic intelligence platform to detect health threats earlier and respond faster .
On artificial intelligence, Dr. Tedros acknowledged unprecedented opportunities - from strengthening surveillance and accelerating research to supporting earlier diagnosis and expanding access to care - but insisted that AI must be safe, ethical, evidence-based, and equitable, governed responsibly with people at the centre . WHO, together with ITU and WIPO, is helping countries adopt AI with confidence through guidance, standards, and governance frameworks, and is working with academia and industry to improve the quality and reliability of AI tools . He delivered what became one of the session's most resonant formulations: "The digital divide is a health divide" , warning that unless gaps in connectivity, capacity, governance, and financing are closed, digital transformation risks widening inequalities rather than reducing them . The future of digital cooperation, he concluded, will be defined not only by technology but by trust, solidarity, and partnership .
##
World Meteorological Organisation - Open Data as the Foundation for Trustworthy AI
WMO Secretary General Celeste Saulo offered a grounding counterpoint to the prevailing enthusiasm for AI, drawing on decades of meteorological experience to argue that AI is only as good as the data behind it . She illustrated this through WMO's AI Challenge for Drought Detection, launched that week in partnership with ITU, UN Habitat, UNCCD, and other partners, in which teams from around the world are using AI to improve early detection of droughts . None of those models, she noted, can succeed without decades of observations from satellites, weather stations, rivers, and oceans, contributed by countries worldwide and exchanged through WMO . AI may generate new insights, but international cooperation provides the knowledge on which those insights depend . Crucially, she argued, AI does not create resilience by itself - it enhances it .
Saulo described how WMO has spent decades building the foundations that trustworthy AI requires, sharing science, knowledge, software, and above all data, based on the conviction that weather data is a global public good . In 2021, WMO members unanimously adopted a unified data policy committing to exchange Earth system data freely and without restriction . From this, WMO set a clear objective: one open system to move the world's weather data in real time around the clock . The resulting WIS2 platform, now connecting more than 100 connection points in over 90 countries on every continent, serves as the open backbone of all forecasting, including AI . She described how 11 countries - including the United States, China, Brazil, Canada, France, Germany, Japan, Morocco, Saudi Arabia, South Korea, and the United Kingdom - operate the global backbone of WIS2, united not by political alignment but by a common language of open standards that allow each country to remain independent while operating as a global network . Whatever may divide these countries elsewhere, she observed, here they build together - pure cooperation through technology for the common good . Her conclusion was unambiguous: trustworthy AI does not begin with algorithms but with open data, standards, open-source software, and the capacity to use them .
##
UN Economic Commission for Europe - Regulatory Frameworks for Digital Innovation
UNECE Executive Secretary Tatiana Molcean reflected on twenty years of WSIS, arguing that the key lesson is that digital transformation is not an end in itself but must be measured by how it impacts daily lives - bringing economic opportunities, better health, better environmental monitoring, and sustainable development . She highlighted UNECE's practical regulatory work, noting that just two weeks prior, member states, the private sector, and the broader community had approved the first regulatory framework for fully driverless automated driving systems . This achievement, she argued, demonstrates that international cooperation on digital innovation is achievable even in challenging times .
Molcean echoed the points made by Dr. Tedros and Saulo on the trustworthiness of data, calling it essential when talking about digital transformation . She stressed the importance of supporting all countries to ensure they have the right skills and capacities to process and benefit from digital solutions, framing this as one of the key ways to address inequality .
##
UNCTAD - Making Connectivity Economically Meaningful
UNCTAD Acting Secretary General Pedro Manuel Moreno reframed the central challenge of digital inclusion with a vivid analogy: the phone in one's pocket can either entertain or employ, and in most of the developing world, it mostly entertains . He noted that some countries rank among the highest in the world for social media use and among the lowest for online sales or internet penetration at the workplace . The same phone that streams a video can take an order, issue an invoice, acquire a customer, and pay a supplier - but whether it does depends on things the user never sees: a payment system that works, a parcel service that delivers, a law that protects the buyer, and the skills to run a business . When one of those links fails, people keep scrolling and income goes elsewhere .
Moreno described UNCTAD's E-Trade for All initiative - which marked its 10th anniversary the day before the panel - now with 35 organisations, providing developing countries a single entry point into the UN system's digital support, with e-trade readiness assessments mapping the full chain of payments, logistics, skills, and legal frameworks in close to 40 developing countries . The E-Trade for Women programme has helped more than 200 women-led digital businesses grow . As chair of UNGIS, UNCTAD also tabled a joint implementation roadmap to align WSIS and the Global Digital Compact . He concluded with a forward-looking formulation: the first 20 years of the summit put much of the world online; the next 10 will be judged by what being online is worth - a sale, a job, an income .
##
UNITAR - Addressing the Capacity Divide
UNITAR Executive Director Michelle Gyles-McDonnough introduced a concept that would resonate throughout the remainder of the session: the capacity divide. Referencing the Independent International Scientific Panel on AI report released that same week, she argued that beyond the digital divide, countries are now confronting a capacity divide - access to technology alone is not enough, and countries need the skills, institutions, leadership, and policy capabilities to harness digital technologies responsibly and effectively . She was explicit that closing this divide must be a strategic priority, not the add-on it is usually treated as .
UNITAR contributes across key WSIS action lines through five areas of action: executive education for decision-makers, academic and professional learning programmes delivered with universities in both the Global North and Global South through its network of 33 CIFAL centres, AI governance training and partnerships across the UN system including through the ITU-led AI Skills Coalition and the newly launched Global AI Capacity Development Network of Centres, tailored advisory support for governments on skills gaps and workforce upskilling, and research and innovation on AI-enabled adult learning . In 2025 alone, UNITAR's learning programmes reach almost 600 learners across 105 countries each year .
##
UNRISD - Social Justice as a Prerequisite, Not an Afterthought
Magdalena Sepulveda of the UN Research Institute for Social Development offered the session's most explicitly critical perspective on the dominant techno-optimist narrative. She stated that at UNRISD, when many discussions on AI speak about technology, they speak about people, asking three simple questions: who benefits, who is left behind, and what policies and institutions are needed to ensure AI promotes sustainable and inclusive development . She argued that the measure of success of AI cannot be determined solely by what AI can do, but must be measured by how AI contributes to societies . UNRISD generates evidence not only about growth and innovation but also about how AI impacts inequality within and between countries, in jobs, in public services, and in trust and well-being . She concluded with a clear normative position: if AI is going to contribute to sustainable development, social justice cannot be an afterthought - it must be included from the very beginning .
##
UN System Staff College - The Last Mile is People
Miguel Panadero of the UN System Staff College described the College's role as turning digital ambition into real institutional capability . Technology can open enormous opportunities, he argued, but only if organisations have the leadership, skills, governance, and partnerships to use it responsibly, inclusively, and in support of the joint mission for sustainable development . The College builds digital capabilities through learning paths and specialised academies, invests in shared digital learning infrastructure - including the Blue Line Learning Marketplace, which connects tens of thousands of UN colleagues worldwide, is open to all including those with least access to in-house formal learning support, and recently won the 2026 Learning Platform of the Year Award - and promotes responsible and human-centred AI through tools such as a De-Biasing Coach that helps colleagues identify bias, question assumptions, and strengthen critical thinking . Together with UNITAR, the College co-leads UNSDG Learn, which offers free learning opportunities and has reached more than 1 million users worldwide . He closed with a formulation that served as a synthesis of the panel's human-centred themes: the last mile of digital transformation is not technology - it is people .
---
#
Transition to Panel Two: From Vision to Implementation
ITU Deputy Secretary General Tomas Lamanauskas opened the second panel by grounding the discussion in sobering statistics: only 35% of SDG targets are on track or making moderate progress, 18% are moving backwards, and 50% are stalled . He described the WSIS process as a uniquely collaborative UN effort, with over 50,000 stakeholders from 160 countries involved and UNGIS bringing together 31 UN entities and 17 observers - nearly 50 entities in total - to coordinate system-wide delivery . He noted that WSIS Plus 20 had assigned UNGIS additional responsibilities, including work on financing for digital development . He framed the second panel's purpose as moving from strategy to delivery, inviting senior officials to share concrete AI and digital initiatives with measurable impact on the ground . Due to other commitments, Lamanauskas handed over moderation of the panel's discussion to his colleague Gitanjali.
---
#
Panel Two: Concrete Implementation Across the UN System
##
ILO - Labour Markets, Research, and Governance
ILO Assistant Director General Beate Andrees, whose organisation serves as the WSIS Action Line lead facilitator for e-employment, highlighted three areas where the ILO sees impact but needs to scale up. She noted that the ILO turned 100 in 2019, and that at that anniversary its constituents - labour ministries, workers' and employers' organisations - adopted a new declaration on the future of work, which serves as the ILO's guiding parameter for its digital work. First, research: the ILO has conducted extensive collaborative research on AI's impact on occupations and is now working with the World Bank on macroeconomic impacts . Second, capacity building: the ILO's International Training Centre in Turin has established programmes for labour ministries and workers' and employers' organisations to manage digital transformations . Third, governance: the ILO has set up a policy tracker for the platform economy and, just the previous month, adopted the first-ever international treaty on decent work in the platform economy .
##
UNESCO - Civil Servant Training and Ethics at Scale
UNESCO Assistant Director General Mariya Gabriel described four concrete areas of implementation. First, UNESCO is building AI and digital governance capacities at scale, with more than 35,000 civil servants already trained through the UNESCO-Oxford Programme on AI and Digital Transformation in Government, and the Spark AI Alliance operating as a network in more than 100 countries . Second, UNESCO supports member states in implementing the Recommendation on the Ethics of AI, adopted by all 193 UNESCO member states, including AI readiness assessments, policy development, and guidance on bias, transparency, and accountability . Third, UNESCO promotes evidence-based digital governance through scalable tools, including a data governance toolkit developed with ITU, UNDP, and the African Union, and the Internet Universality Indicators . Fourth, UNESCO strengthens trust in the digital environment through media information literacy and guidelines for the governance of digital platforms . She emphasised that these initiatives are designed not as standalone programmes but as enablers of a whole-of-UN approach to digital transformation .
##
UN Ops - Shared Infrastructure and Operational Inclusion
UN Ops Assistant Secretary General Sonja Leighton-Kone argued that digital inclusion is not just about connectivity - true inclusion requires foundational public infrastructure, institutional capability, and data governance that makes connectivity useful once it is there . She described three operational pathways: system-wide UN digital infrastructure, including the UN Global Marketplace and the UN WebI Plus platform, which has delivered over $1 billion in goods and services across 128 countries to support UN agencies, financial institutions, and NGOs ; multi-tenant operability, whereby instead of 40 separate agencies building 40 separate tools, UNOPS designs next-generation systems - including end-to-end recruitment and talent management, grant management systems covering everything from awards to final disbursement, and digital procurement systems - with multi-tenant architecture for shared use across sister UN agencies ; and targeted field inclusion, bringing digital transformation and emergency connectivity directly to post-conflict least developed countries and small island developing states . She stressed that the real risk is that the digital agenda moves fastest for those already connected, leaving the hardest to reach further behind , and that the role of UNOPS is to build capability, not dependency .
##
UNDP - Digital as a Cross-Cutting Accelerator
UNDP Chief Digital Officer Robert Opp described how UNDP has hard-coded digital and AI transformation into its strategic plan for 2026 to 2029 as a cross-cutting accelerator across all parts of its business . This involves embedding digital approaches in thematic programmes as well as providing horizontal support on policies and strategies around digital governance, technology foundations including digital public infrastructure, and capacity building in over 80 countries with civil servants, civil society, and community groups . He noted collaboration with ITU and UNESCO on connecting health system digital work with UNDP's cross-cutting support , and highlighted UNDP's work with ITU on one of the UN80 work streams around expertise on demand, for which a digital platform is being built to connect supply and demand for services across the system .
##
UNICEF - Equity, Children, and AI Benchmarks
UNICEF Global Chief Information Officer Kaan Cetinturk argued that scale and impact must be measured not only by numbers but also through the lens of equity, and that digital transformation should empower the most vulnerable - in UNICEF's case, children - and bridge rather than widen the digital divide . He presented two AI initiatives. The first, Ahead of the Storm, is designed to improve preparedness for climate emergencies affecting children: nearly half of the world's children, approximately 1.1 billion, are exposed to multiple climate hazards, and the initiative uses AI to convert complex hazard forecasts and contextual vulnerability data into timely, decision-ready operational data that countries can use to prepare and trigger early action . The second is an early-stage project to develop a child-centric AI benchmark framework, addressing the gap that most AI benchmarks are designed for adults despite children comprising one third of internet users worldwide .
##
UN OHCHR - Human Rights as the Operating System for AI
Anwar Mahfoudh of UN OHCHR described how the human rights community is reimagining how to advance human rights in the digital age, moving beyond traditional and manual approaches to identify where AI can genuinely add value while managing risks and ensuring human rights safeguards are built in . At AI for Good that week, OHCHR launched the Human Rights Data Exchange, a dedicated authoritative platform that brings together scattered data from different UN entities, mechanisms, and institutions, unlocking information that had long sat in the public sphere and making it accessible and interoperable across the entire human rights ecosystem . He also described scaling capabilities on crisis monitoring, including remote monitoring in resource-constrained environments . His central reframing was striking: human rights are not an obstacle to AI but the operating system that AI should be built upon safely, and when built responsibly, AI can strengthen protection, accountability, and scale trusted decision-making across the UN .
##
UN Joint Staff Pension Fund - Pioneering Responsible AI Certification
UN Pension Fund Chief Information Officer Dino Cataldo Dell'Accio described how the Pension Fund has demonstrated that it is possible to use emerging technologies, including AI, in a responsible and ethical manner . In 2021, with technical support from UNICC, the Fund implemented a digital identity solution for Proof of Life, enabling 80,000 retirees residing in more than 192 countries to confirm that they are still alive - replacing a process that had been paper-based for 70 years and raised concerns about fraud and inefficiency . This application gave birth to the United Nations Digital Identity Solution, now joined by 16 UN entities, demonstrating collaboration, interoperability, and shared efficiencies in human and financial resources . The Fund also became a pioneer by achieving the first ISO 42001 certificate on the responsible use of AI within the UN system, demonstrating in a transparent manner how AI has been implemented .
##
UNICC - Shared Digital and AI Infrastructure for the UN System
UNICC Deputy Director Milena Grecuccio argued that the UN system's adoption and development of technology must embed technical, architectural governance, and human rights principles by design, consistently across the system in a transparent and trusted manner . She described UNICC as having built and continuing to build the shared digital foundation for the UN system over 55 years of running UN data centres, including shared digital identity systems, treasury platforms under the leadership of UNHCR that move funds securely across borders to reach beneficiaries no matter how remote or unbanked, and a modern private cloud keeping data secure and protected by UN privileges and immunities . Looking ahead, she outlined a three-pillar framework for extending this shared approach to AI: an AI factory to move solutions safely from experimentation to deployment at scale; an AI gateway to govern transparently how AI is adopted, including legal vetting; and an AI studio to build production-grade AI agents with guardrails embedded .
##
UNICRI - AI for Child Protection and Law Enforcement
Irakli Beridze of UNICRI's Centre for AI and Robotics described how UNICRI has been an early adopter in the area of AI, starting programmes in 2015 to 2017 and opening its Centre for Artificial Intelligence and Robotics in The Hague . UNICRI supports UN member states in understanding how AI can tackle concrete crime areas and how to ensure AI is used by law enforcement in a responsible and human rights-compliant manner . A flagship initiative, AI for Safe Children, undertaken in partnership with the Ministry of Interior of the United Arab Emirates, built an AI global hub providing over 100 AI tools to representatives of 129 countries to tackle the sexual exploitation and abuse of children online . Through this project, UNICRI has trained over 3,500 law enforcement officers from 60 countries, and supported operational undertakings that resulted in saving over 100 children, arresting over 250 perpetrators, and delivering concrete operational support . UNICRI has also recently launched a project together with the OpenAI Foundation on building curriculum on children's safety online.
##
UNIDO - Industrial AI for the Real Economy
UNIDO Chief of Innovation Jason Slater described the Global Alliance on AI for Industry Manufacturing (AIM Global), which brings together governments, industry, and academia to help countries adopt industrial AI responsibly in the real economy, especially in the Global South . The initiative has built a partnership base of around 500 in over 70 countries, with Centres of Excellence spanning Africa, Asia, Europe, Latin America, and the Caribbean . These centres are designed to demonstrate, adapt, and apply industrial AI applications, build local skills and capacity particularly for SMEs, and provide a trusted space with clear guardrails to solve complex challenges . UNIDO is also rolling out an AI and digital economy industry navigator, a country-level diagnostic readiness tool, a governance framework for industrial AI, and an AI for Industry Skills curriculum . He highlighted UNIDO's partnership with ITU on Partner Connect and the Skills Coalition, connecting ITU's work on connectivity and skills with UNIDO's focus on industrial applications . His closing message was direct: responsible AI must reach factories, firms, and workers, not only laboratories and platforms .
---
#
Closing Reflections and Cross-Cutting Themes
In her closing remarks following the first panel, Bogdan-Martin drew out the common threads running across all interventions: the impact of digital transformation on trade, postal services, climate prediction and early warning, health, transport standards, mobility, and the empowerment of women . She identified trust as a recurring theme - not just trust in systems and services or in data, but trust amongst the UN agencies themselves - arguing that this is how the system can succeed, ensuring it is not duplicating or competing but working together to deliver better for member states and the multi-stakeholder community .
Gitanjali, who moderated the second panel, concluded by affirming that the WSIS process is the UN in action, with several UN agencies implementing their work on the ground to have an impact on people all over the world . The session closed with a call for a group photograph as a collective memory of the UN family's unity at the WSIS Forum .
---
#
Synthesis: Agreements, Tensions, and Unresolved Questions
Across both panels, a remarkably high degree of consensus emerged around several foundational principles. All speakers agreed that the digital divide extends beyond connectivity to encompass capacity, skills, governance, and financing gaps . All agreed that trustworthy AI requires high-quality, open, and well-governed data as its foundation . All agreed that people must be placed at the centre of digital transformation and AI governance, with social justice and equity embedded from the outset rather than retrofitted . All agreed that capacity building for policymakers, civil servants, and institutions is a strategic priority, not an add-on . And all agreed that UN system-wide cooperation through shared infrastructure and UNGIS is essential to avoid duplication and deliver effectively for member states .
Beneath this surface consensus, however, meaningful tensions emerged. Bogdan-Martin framed the central challenge primarily as one of connectivity, citing 2.2 billion people still offline , while Moreno argued that the more pressing challenge is what being online is actually worth economically , and Leighton-Kone insisted that true inclusion requires foundational infrastructure and data governance beyond mere connectivity . Saulo explicitly cautioned against treating AI as a primary driver, arguing it merely enhances pre-existing resilience , while Opp positioned AI as a hard-coded strategic accelerator and Slater described it as a clear driver of inclusive industrial development . Gyles-McDonnough introduced the capacity divide as potentially more urgent than the digital divide , while Bogdan-Martin's framing remained anchored in connectivity . And Sepulveda and Mahfoudh argued that social justice and human rights must be constitutive of AI governance from the outset , while Gabriel's approach framed ethics more procedurally through readiness assessments and policy development tools .
Several issues remained unresolved at the close of the session: how to finance digital development at the scale required for least developed countries and small island developing states ; how to prevent digital transformation from widening inequalities in contexts where enabling conditions are absent ; how to ensure AI benchmarks and governance frameworks account for children and other vulnerable populations ; how to achieve genuine interoperability across the forty-plus UN agencies that risk building separate, duplicative digital tools ; and how to ensure that the rapid deployment of AI technologies keeps pace with the protection of civic space and human rights, particularly in conflict-affected environments . These unresolved questions, taken together, define the agenda for the renewed WSIS mandate to 2035 and the broader UN80 reform process that the session sought to advance.
UN family leads WSIS action lines and must carry forward the renewed WSIS mandate to 2035 through multi-stakeholder cooperation - Opening remarks on UN unity (Doreen Bogdan-Martin)
Arg. 1Doreen Bogdan-Martin emphasises that the UN family collectively leads as facilitators of the WSIS action lines and bears a shared responsibility to advance the WSIS vision through to 2035. She highlights that 20 years of multi-stakeholder cooperation has proven effective, bringing together governments, civil society, academia, the private sector, and international organisations to drive progress towards a better digital future.
She noted that it is the UN family that leads as facilitators of the WSIS action lines and forms UNGIS , and that it is their collective responsibility to carry the WSIS vision and renewed mandate to 2035 . She further stated that the past 20 years has proven that multi-stakeholder cooperation works, building a time-tested platform involving governments, civil society, academia, the private sector, international organisations, and the UN system .
on: The WSIS mandate renewal to 2035 and multi-stakeholder cooperation remain essential frameworks for advancing digital transformation and sustainable development
With 2.2 billion people still offline, the digital divide remains a critical challenge requiring targeted action across the UN system - Scale of the digital divide (Doreen Bogdan-Martin)
Arg. 2Doreen Bogdan-Martin underscores that despite progress, the work of the UN system is far from complete given that 2.2 billion people remain without internet access. This statistic frames the urgency of continued and coordinated action across all UN agencies to address the digital divide.
She explicitly stated that with 2.2 billion people still offline, the work is far from over , using this figure to contextualise the need for panellists to share their thoughts on carrying forward the WSIS mandate and strengthening system-wide cooperation .
on: The digital divide extends beyond connectivity to encompass capacity, skills, governance, and financing gaps
The UN system must avoid duplication and competition, working together through UNGIS and shared platforms to deliver better for member states and the multi-stakeholder community - UN system cooperation imperative (Doreen Bogdan-Martin)
Arg. 3Doreen Bogdan-Martin stresses that trust — not only in systems and data, but amongst UN agencies themselves — is fundamental to the system's success. She argues that the UN must ensure it is not duplicating or competing, but genuinely working together to deliver better outcomes for member states and the broader multi-stakeholder community.
In her closing remarks, she noted that there were many common references to trust across panellists' interventions, and that this trust must extend to trust amongst the UN agencies themselves . She explicitly stated that the system can succeed by ensuring they are not duplicating or competing, but working together to deliver better for member states and the multi-stakeholder community .
on: UN system-wide cooperation, shared infrastructure, and avoiding duplication are essential for effective digital transformation delivery
WSIS process is a uniquely collaborative UN effort involving over 50,000 stakeholders from 160 countries, with UNGIS coordinating system-wide delivery - UN system coordination (Tomas Lamanauskas)
Arg. 1Tomas Lamanauskas describes the WSIS process as a genuinely unified UN effort, with UNGIS bringing together 31 UN entities and 17 observers to coordinate delivery across the system. He highlights the scale of stakeholder engagement and notes that UNGIS has been recognised by member states as an important coordination platform, with additional responsibilities assigned under WSIS Plus 20.
He noted that directly into the WSIS process, over 50,000 stakeholders from 160 countries were involved , and that UNGIS, established in 2006, brings together 31 UN entities and 17 observers . He also mentioned that UNGIS has been recognised by member states as an important platform and has been given additional work under the WSIS Plus 20 review, including on financing for digital development .
on: The WSIS mandate renewal to 2035 and multi-stakeholder cooperation remain essential frameworks for advancing digital transformation and sustainable development
UNGIS brings together 31 UN entities and 17 observers to coordinate system-wide digital delivery, and has been tasked with additional work under WSIS Plus 20, including on financing for digital development - UNGIS as coordination mechanism (Tomas Lamanauskas)
Arg. 2Lamanauskas explains that UNGIS serves as the primary coordination mechanism for the UN system's digital work, uniting nearly 50 entities to ensure coherent delivery. Under the WSIS Plus 20 review, UNGIS has been assigned new responsibilities, including examining financing options and gaps for digital development.
He stated that UNGIS was established in 2006 and brings together 31 UN entities and 17 observers, making it nearly 50 entities working together . He noted that in the WSIS Plus 20 review, UNGIS received additional work including on financing for digital development, looking at options, gaps, and implications .
Digital cooperation remains essential to sustainable development, and WHO's role is to ensure digital health transformation is trusted, equitable, and responsive to people's needs - Digital cooperation for health (Dr. Tedros Adhanom Ghebreyesus)
Arg. 1Dr. Tedros argues that the renewal of the WSIS mandate to 2035 signals the continued importance of digital cooperation for sustainable development. For WHO, the central challenge is not whether technology will transform health — it already is — but ensuring that transformation is trusted, equitable, and people-centred.
He stated that the renewal of the WSIS mandate sends an important message that digital cooperation remains essential to achieving sustainable development . He noted that for WHO, the real challenge is ensuring that digital transformation is trusted, equitable, and responsive to people's needs , and that this principle guides WHO's work as co-facilitator of the WSIS action line on e-health .
The digital divide is also a health divide; unless gaps in connectivity, capacity, governance, and financing are closed, digital transformation risks widening inequalities - Digital divide as health divide (Dr. Tedros Adhanom Ghebreyesus)
Arg. 2Dr. Tedros makes the powerful argument that the digital divide and the health divide are inseparable — failure to close gaps in connectivity, capacity, governance, and financing means digital transformation could deepen rather than reduce inequalities. He calls for trust, solidarity, and partnership as the defining features of future digital cooperation.
He explicitly stated that the digital divide is a health divide , and that unless gaps in connectivity, capacity, governance, and financing are closed, digital transformation risks widening inequalities rather than reducing them . He concluded that the future of digital cooperation will be defined not only by technology, but by trust, solidarity, and partnership .
AI is creating unprecedented opportunities for health, including strengthening surveillance, accelerating research, supporting earlier diagnosis, and expanding access to care, but must be safe, ethical, and equitable - AI opportunities in health (Dr. Tedros Adhanom Ghebreyesus)
Arg. 3Dr. Tedros acknowledges the transformative potential of AI in health, citing a wide range of applications from disease surveillance to drug development. However, he insists that AI must be governed responsibly, with people at the centre, and that WHO is working with ITU and WIPO to help countries adopt AI safely through guidance, standards, and governance frameworks.
He listed specific AI applications in health including strengthening surveillance, accelerating research, supporting earlier diagnosis, assisting health workers, advancing drug and vaccine development, and expanding access to care . He emphasised that AI must be safe, ethical, evidence-based, and equitable, and must be governed responsibly with people at the centre . He noted that WHO, together with ITU and WIPO, is helping countries adopt AI with confidence through guidance, standards, and governance frameworks , and referenced a joint publication on AI-enabled health innovation and IP .
AI must be governed responsibly with people at the centre; WHO, ITU, and WIPO are helping countries adopt AI with confidence through guidance, standards, and governance frameworks - Responsible AI governance frameworks (Dr. Tedros Adhanom Ghebreyesus)
Arg. 4Dr. Tedros argues that innovation alone is insufficient and that AI governance must place people at the centre. He highlights the collaborative work of WHO, ITU, and WIPO in providing countries with the tools they need to adopt AI responsibly, including guidance, standards, and governance frameworks.
He stated that AI must be safe, ethical, evidence-based, and equitable, and must be governed responsibly with people at the centre . He noted that WHO, together with ITU and WIPO, is helping countries adopt AI with confidence through guidance, standards, and governance frameworks , and that they are also working with academia and industry to improve the quality and reliability of AI tools and strengthen countries' capacity to use them effectively .
on: AI must be governed responsibly, ethically, and with human rights safeguards built in from the outset
Post offices serve as inclusive digital access points, especially in rural areas, and UPU is working to connect the estimated 100,000 post offices that still lack meaningful internet connection - Postal sector as digital inclusion enabler (Masahiko Metoki)
Arg. 1Masahiko Metoki argues that the postal sector plays a unique and underappreciated role in digital inclusion, particularly in rural areas where post offices serve as accessible points for digital services. He highlights that the WSIS Plus 20 document explicitly recognises post offices as inclusive access points, and that UPU is committed to closing the connectivity gap for the estimated 100,000 post offices still without meaningful internet.
He noted that the WSIS Plus 20 document approved by the UN General Assembly clearly states that post offices are inclusive access points for delivering digital services in rural areas . He cited approximately 600,000 post offices worldwide, most in rural areas , and estimated that around 100,000 post offices do not have meaningful internet connection . He stated that UPU will continue to connect communities through the post office, both through its own projects and in partnership with other UN organisations .
Post offices, with around 600,000 locations worldwide mostly in rural areas, can provide digital services cost-effectively to rural communities, elderly people, and women through initiatives like Post for Women - Post offices bridging rural digital divide (Masahiko Metoki)
Arg. 2Metoki makes the case that the physical reach of the postal network — with hundreds of thousands of locations predominantly in rural areas — gives the postal sector a special role in bridging the digital divide for underserved populations. He highlights specific initiatives such as Post for Women, which empowers women-led businesses in rural areas to build digital skills and sell products online.
He stated that there are around 600,000 post offices worldwide, most in rural areas, giving the postal sector a special role . He noted that post offices can provide digital services cost-effectively to people who may not easily access the internet, smartphones, or digital skills , and that many postal operators now offer e-commerce, e-governance, or digital financial services in person at post office counters . He specifically mentioned the Post for Women initiative, which empowers women-led business enterprises in rural areas to build digital skills and sell products online .
UNCTAD's E-Trade for All initiative, now with 35 organisations, provides developing countries a single entry point into the UN system's digital support, and the next decade will be judged by what being online is worth economically - Digital trade and economic inclusion (Pedro Manuel Moreno)
Arg. 1Pedro Manuel Moreno argues that UNCTAD's role is to ensure that digital infrastructure built by governments and the private sector reaches everyone, particularly through the economic lens. He highlights the E-Trade for All initiative as a concrete mechanism for coordinating UN digital support to developing countries, and frames the next decade of WSIS as one that must be judged by the economic value generated for people online.
He noted that UNCTAD launched the E-Trade for All initiative ten years ago, now with 35 organisations, offering developing countries a single door into the system's digital support . He stated that E-Trade readiness assessments have mapped the full chain of payments, logistics, skills, and legal frameworks in close to 40 developing countries , and that E-Trade for Women has helped more than 200 women-led digital businesses grow . He concluded that the first 20 years of the summit put much of the world online, and the next 10 will be judged by what being online is worth - a sale, a job, an income .
on: Whether the primary digital challenge is connectivity/access or economic utility of existing connectivity
In most of the developing world, mobile phones are used predominantly for entertainment rather than economic activity; closing this gap requires functional payment systems, logistics, legal frameworks, and skills - Phone as economic tool vs. entertainment device (Pedro Manuel Moreno)
Arg. 2Moreno highlights a critical paradox in digital development: even in countries with high social media use, the same devices are rarely used for economic activity such as online sales or business transactions. He argues that closing this gap requires not just connectivity but a full ecosystem of enabling conditions including payment systems, logistics, legal protections, and business skills.
He noted that in most of the developing world, the phone mostly entertains rather than employs, with some countries ranking among the highest in the world for social media use but among the lowest for online sales or internet penetration at the workplace . He explained that whether a phone is used for economic activity depends on things the user never sees - a payment system that works, a parcel service that delivers, a law that protects the buyer, and the skills to run a business . He stated that when one of those links fails, people keep scrolling and the income goes elsewhere .
UNCTAD, as chair of UNGIS, tabled a joint implementation roadmap to align WSIS and the Global Digital Compact, demonstrating concrete system-wide coordination - Aligning WSIS and Global Digital Compact (Pedro Manuel Moreno)
Arg. 3Moreno highlights UNCTAD's role as chair of UNGIS in tabling a joint implementation roadmap to align the WSIS process with the Global Digital Compact, representing a concrete example of system-wide coordination. This initiative demonstrates how the UN system is working to ensure coherence between its major digital governance frameworks.
He stated that as chair of the UN Group on the Information Society, UNCTAD tabled a joint implementation roadmap to align WSIS and the Global Digital Compact .
on: UN system-wide cooperation, shared infrastructure, and avoiding duplication are essential for effective digital transformation delivery
UNECE demonstrates digital cooperation through practical regulatory outcomes, such as the first regulatory framework for fully driverless automated driving systems, approved just weeks prior - Regulatory frameworks for digital innovation (Tatiana Molcean)
Arg. 1Tatiana Molcean argues that digital transformation is not an end in itself but must be measured by its impact on daily lives and economic opportunities. She uses UNECE's work on automated vehicles as a concrete example of how digital innovation, combined with international regulatory cooperation, can deliver tangible outcomes even in challenging times.
She noted that UNECE works on regulations for cars, transport, and vehicles, and that two weeks prior to the forum, member states, the community, and the private sector came together and approved the first regulatory framework for fully driverless automated driving systems . She presented this as evidence that international cooperation on digital innovation can deliver results .
Trustworthiness of data is essential for digital transformation, and all countries must have the right skills and capacities to process and benefit from digital solutions - Data trustworthiness and national capacity (Tatiana Molcean)
Arg. 2Molcean echoes points made by other panellists about the critical importance of data trustworthiness, arguing that it is essential for meaningful digital transformation. She stresses that ensuring all countries have the right skills and capacities to process and benefit from digital solutions is a key pathway to addressing inequality.
She joined the points made by Dr. Tedros and Celeste Saulo about the trustworthiness of data, calling it essential when talking about digital transformation . She stated that attention must be paid to supporting all countries to ensure they have the right skills and capacities to process and benefit from digital solutions , and expressed hope that this would be a key way to solve inequality and ensure the benefits of digital transformation are shared .
on: Whether the digital divide or the capacity divide is the more critical challenge to address
AI is only as good as the data behind it; WMO's decades of open Earth system data sharing through its unified data policy and WIS2 platform provide the trusted foundation that AI in meteorology requires - Data foundations for trustworthy AI (Celeste Saulo)
Arg. 1Celeste Saulo argues that the excitement around AI must be tempered by recognition that AI is only as good as the data underpinning it. She uses WMO's experience in meteorology to demonstrate that decades of international data sharing, codified through a unified data policy and operationalised through the WIS2 platform, are what make trustworthy AI applications possible.
She stated that in meteorology, a simple lesson has been learned: AI is only as good as the data behind it . She noted that AI models for drought detection cannot succeed without decades of observations from satellites, weather stations, rivers, and oceans, contributed by countries around the world and exchanged through WMO . She described WMO's 2021 unified data policy, unanimously adopted by members, committing to exchange Earth system data freely and without restriction , and the WIS2 platform as the open successor to ageing systems, serving as the backbone of all forecasting including AI .
on: Whether AI is the primary driver of digital transformation or merely an enhancer dependent on pre-existing data and cooperation foundations
WMO's 2021 unified data policy commits members to exchange Earth system data freely and without restriction, establishing weather data as a global public good and underpinning trustworthy AI in meteorology - Open data as a global public good (Celeste Saulo)
Arg. 2Saulo presents WMO's unified data policy as a landmark example of international agreement to treat data as a global public good. She argues that this commitment to free and open data exchange, operationalised through the WIS2 platform with over 100 connection points in more than 90 countries, is the foundation upon which trustworthy AI in meteorology is built.
She stated that WMO holds a simple conviction: weather data is a global public good, and so in many cases is the software itself . She noted that in 2021, WMO members unanimously adopted the unified data policy, committing to exchange Earth system data freely and without restriction . She described WIS2 as having more than 100 connection points in over 90 countries on every continent, from small island states in the Caribbean and the Pacific to nations across Africa and Latin America .
Beyond the digital divide, there is now a capacity divide; access to technology alone is insufficient without the skills, institutions, leadership, and policy capabilities to harness it responsibly - The capacity divide (Michelle Gyles-McDonnough)
Arg. 1Michelle Gyles-McDonnough introduces the concept of the 'capacity divide' as a challenge that goes beyond mere connectivity. She argues that even when countries have access to technology, they often lack the skills, institutional frameworks, leadership, and policy capabilities needed to harness digital technologies responsibly and effectively.
She referenced the Independent International Scientific Panel on AI report released that week, which highlights a growing challenge: beyond the digital divide, countries are now confronting the capacity divide . She stated that access to technology alone is not enough, and that countries need the skills, institutions, leadership, and policy capabilities to harness digital technologies responsibly and effectively .
on: Whether the digital divide or the capacity divide is the more critical challenge to address
UNITAR strengthens the capabilities of policymakers, diplomats, and public servants through executive education, AI governance training, and tailored advisory support, reaching almost 600,000 learners across 105 countries annually - Capacity building for policymakers (Michelle Gyles-McDonnough)
Arg. 2Gyles-McDonnough outlines UNITAR's specific contribution to closing the capacity divide through five areas of action targeting decision-makers and public servants. She emphasises the scale and reach of UNITAR's programmes, which span executive education, AI governance training, and tailored advisory support across more than 100 countries.
She described UNITAR's five areas of action including executive education for decision-makers, academic and professional learning programmes delivered through a network of 33 CIFAL centres, AI governance training and partnerships across the UN system including collaboration with the ITU-led AI Skills Coalition , and tailored advisory support including AI training programmes for youth and women in Malaysia and Ukraine . She stated that in 2025 alone, UNITAR's learning programmes reach almost 600,000 learners across 105 countries each year .
UNRISD focuses on the social dimensions of AI, asking who benefits, who is left behind, and what policies are needed; social justice must be included from the very beginning of AI governance, not as an afterthought - Social justice in AI governance (Magdalena Sepulveda)
Arg. 1Magdalena Sepulveda argues that UNRISD's distinctive contribution is to place people — rather than technology — at the centre of AI discussions, asking fundamental questions about who benefits and who is left behind. She insists that social justice cannot be an afterthought in AI governance but must be embedded from the very beginning.
She stated that at UNRISD, when many discussions on AI speak about technology, they speak about people, asking three simple questions: who benefits, who is left behind, and what policies and institutions are needed to ensure AI promotes sustainable and inclusive development . She noted that UNRISD generates evidence not only about growth and innovation but also about how AI impacts inequality within and between countries, in jobs, public services, trust, and well-being . She concluded that if AI is going to contribute to sustainable development, social justice cannot be an afterthought but must be included from the very beginning .
on: Whether social justice and human rights should be foundational to AI governance from the outset or integrated as part of a broader governance process
The UN System Staff College helps UN personnel build digital capabilities through learning programmes, shared digital learning infrastructure such as the Blue Line Learning Marketplace, and responsible AI tools including a De-Biasing Coach - Building digital capabilities within the UN system (Miguel Panadero)
Arg. 1Miguel Panadero describes the UN System Staff College's role in turning digital ambition into real institutional capability within the UN system itself. He outlines three practical ways the College supports this: building digital capabilities through learning programmes, investing in shared digital learning infrastructure, and promoting responsible and human-centred AI.
He described the digital for sustainability learning path and specialised academies that help colleagues understand how emerging technologies work, where they create public value, and how they can be governed responsibly . He noted the Blue Line Learning Marketplace, which connects tens of thousands of UN colleagues worldwide and is open to all, including national professionals and locally recruited personnel . He also described the De-Biasing Coach, an AI-enabled learning assistant that helps colleagues identify bias, question assumptions, and strengthen critical thinking . Together with UNITAR, the College co-leads UNSDG Learn, which has reached more than 1 million users worldwide .
ILO conducts collaborative research on AI's impact on occupations and labour markets, runs capacity-building programmes for labour ministries and workers' organisations, and has adopted the first international treaty on decent work in the platform economy - Labour market capacity building and governance (Beate Andrees)
Arg. 1Beate Andrees outlines ILO's three-pronged approach to managing the digital transformation of work: research on AI's impact on occupations, capacity building for labour ministries and workers' organisations, and governance through international standards. She highlights the adoption of the first-ever international treaty on decent work in the platform economy as a landmark achievement.
She described ILO's collaborative research on the exposure of AI on different occupations and macroeconomic impacts, conducted in collaboration with the World Bank . She noted the International Training Centre in Turin, which has set up capacity-building programmes for labour ministries, workers' and employers' organisations . She highlighted that last month, ILO adopted the first-ever international treaty on decent work in the platform economy , and that ILO has set up a policy tracker for policies on the platform economy .
UNESCO is training over 35,000 civil servants through the UNESCO-Oxford Programme on AI and Digital Transformation in Government, and operates the Spark AI Alliance across more than 100 countries - Civil servant AI training at scale (Mariya Gabriel)
Arg. 1Mariya Gabriel highlights UNESCO's concrete efforts to build AI and digital governance capacities at scale, with a particular focus on civil servants and public administration. She presents two flagship initiatives — the UNESCO-Oxford Programme and the Spark AI Alliance — as practical tools for equipping public servants with AI governance skills.
She stated that more than 35,000 civil servants are already being trained through the UNESCO-Oxford Programme on AI and Digital Transformation in Government . She also described the Spark AI Alliance as a network operating in more than 100 countries, providing practical skills for AI governance, public sector modernisation, and digital transformation .
UNESCO supports member states in implementing its Recommendation on the Ethics of AI, adopted by all 193 member states, including AI readiness assessments, policy development, and guidance on bias, transparency, and accountability - Ethics of AI implementation (Mariya Gabriel)
Arg. 2Gabriel argues that UNESCO's Recommendation on the Ethics of AI, adopted by all 193 member states, provides a universal framework for responsible AI governance. She describes the practical support UNESCO offers to member states in implementing this recommendation, including readiness assessments, policy development, and guidance on addressing risks such as bias and lack of transparency.
She noted that UNESCO supports member states in implementing the Recommendation on the Ethics of AI, which was adopted by all 193 UNESCO member states . This support includes AI readiness assessments, policy development, institutional safeguards, and guidance on how to address risks such as bias, transparency, and accountability .
on: Whether social justice and human rights should be foundational to AI governance from the outset or integrated as part of a broader governance process
UNOPS manages the UN Global Marketplace and multi-tenant architecture systems for recruitment, grant management, and procurement, enabling shared interoperable approaches across 40-plus UN agencies rather than duplicating separate tools - Multi-tenant shared operational systems (Sonja Leighton-Kone)
Arg. 1Sonja Leighton-Kone argues that UNOPS contributes to UN system-wide efficiency by designing shared, multi-tenant digital systems that multiple agencies can use, rather than each building separate tools. This approach promotes interoperability and cost-efficiency across the UN system.
She described UNOPS managing the UN Global Marketplace as a common entry point for vendors across the entire UN system, and running platforms like UN WebI Plus, which has delivered over $1 billion in goods and services across 128 countries . She explained that instead of 40 separate agencies building 40 separate tools, UNOPS designs next-generation systems with multi-tenant architecture, including end-to-end recruitment and talent management, grant management, and digital procurement systems made available to sister UN agencies .
UNOPS brings digital transformation and emergency connectivity directly to post-conflict least developed countries and small island developing states, focusing on building capability rather than dependency - Digital inclusion for hardest-to-reach communities (Sonja Leighton-Kone)
Arg. 2Leighton-Kone highlights the risk that the digital agenda moves fastest for those already connected, leaving the hardest-to-reach communities further behind. She argues that UNOPS addresses this by bringing digital transformation and emergency connectivity directly to post-conflict LDCs and small island developing states, with a focus on building local capability rather than creating dependency on external providers.
She stated that the real risk is that the digital agenda moves fastest for those already connected, leaving the hardest to reach further behind . She noted that UNOPS brings digital transformation and emergency connectivity directly to post-conflict LDCs and small island development states , and that their role is to build capability, not dependency on external providers, including by designing procurement frameworks with safety and data governance requirements built in from the start .
on: Whether the primary digital challenge is connectivity/access or economic utility of existing connectivity
UNDP has hard-coded digital and AI transformation into its 2026–2029 strategic plan as a cross-cutting accelerator, supporting both horizontal digital governance and vertical thematic programme integration - AI and digital as strategic accelerators for development (Robert Opp)
Arg. 1Robert Opp explains that UNDP has embedded digital and AI transformation as a strategic accelerator across all parts of its business in its 2026–2029 strategic plan. This involves both horizontal support for cross-cutting digital elements such as policies, strategies, and digital public infrastructure, and vertical integration of digital approaches within specific thematic programmes.
He stated that UNDP has hard-coded digital and AI transformations into its strategic plan for 2026 to 2029 as an accelerator that goes across all parts of its business . He described this as involving both horizontal support for digital governance, technology foundations, and capacity building, and vertical integration within UNDP's many thematic programmes . He also noted UNDP's work with ITU and leadership on one of the UN80 work streams around expertise on demand, powered by digital approaches, and a digital platform being built to connect services and available offers across the system .
on: Whether AI is the primary driver of digital transformation or merely an enhancer dependent on pre-existing data and cooperation foundations
UNDP conducts capacity building in over 80 countries with civil servants, civil society, and community groups, embedding digital approaches both horizontally across societies and vertically within thematic programmes - Broad-based digital capacity building (Robert Opp)
Arg. 2Opp highlights UNDP's extensive capacity-building work as a key component of its digital transformation approach, reaching civil servants, civil society, and community groups across more than 80 countries. He frames this as part of a broader horizontal and vertical strategy that mirrors how national digital transformations are evolving.
He stated that UNDP conducts capacity building in over 80 countries with civil servants, but also civil society and many different community groups and beyond . He described this as part of a horizontal and vertical framing that is also reflected in national digital and AI transformations, including across ministries and cross-cutting elements like digital public infrastructure .
UNICEF's digital transformation initiatives must be measured through the lens of equity, ensuring they empower the most vulnerable, particularly children, and bridge rather than widen the digital divide - Equity lens for digital transformation (Kaan Cetinturk)
Arg. 1Kaan Cetinturk argues that as the conversation shifts from proof-of-concept to scale and impact, the measure of success must go beyond numbers to include equity. He insists that digital transformation should empower the most vulnerable — in UNICEF's case, children — and must actively bridge rather than widen the digital divide.
He stated that scale and impact must be measured not only by numbers but also through the lens of equity . He argued that digital transformation should empower the most vulnerable, in UNICEF's case children, and expand access and opportunity to help bridge the digital divide rather than making it wider .
on: Whether AI is the primary driver of digital transformation or merely an enhancer dependent on pre-existing data and cooperation foundations
UNICEF's Ahead of the Storm initiative uses AI to convert complex climate hazard forecasts into decision-ready operational data, helping countries prepare for climate emergencies affecting 1.1 billion children - AI for climate emergency preparedness for children (Kaan Cetinturk)
Arg. 2Cetinturk presents UNICEF's Ahead of the Storm initiative as a concrete example of AI being used to protect children from climate emergencies. The initiative converts complex hazard forecasts and vulnerability data into actionable operational information that countries can use to prepare and trigger early action.
He described the Ahead of the Storm initiative as designed to improve preparedness for climate emergencies affecting children, noting that nearly half of the world's children - approximately 1.1 billion - are exposed to multiple climate hazards including floods, storms, and extreme heat . He explained that AI is used to convert uncertain complex hazard forecasts and contextual vulnerability data into timely, decision-ready operational data that countries can use to prepare and trigger early action, including estimating how many children may be affected and where essential services could be disrupted .
Human rights are not an obstacle to AI but the operating system upon which AI should be built; OHCHR launched the Human Rights Data Exchange to make scattered human rights data accessible and interoperable across the UN ecosystem - Human rights as the foundation for AI (Anwar Mahfoudh)
Arg. 1Anwar Mahfoudh reframes the relationship between human rights and AI, arguing that human rights are not a constraint on AI but rather the foundational operating system upon which AI should be built. He presents OHCHR's Human Rights Data Exchange as a flagship initiative that demonstrates how AI can strengthen human rights protection when built responsibly.
He stated that human rights are not an obstacle to AI but the operating system that AI should be built upon safely, and that when built responsibly, AI can strengthen protection, accountability, and scale trusted decision-making across the UN . He described the Human Rights Data Exchange, launched at AI for Good, as a dedicated authoritative platform that brings together scattered data from different UN entities, mechanisms, and institutions, powered by a generative AI system, making human rights information accessible, connected, and interoperable . He also mentioned the Human Rights Council MAPA, which feeds into the mandate registry of UN80 .
on: Whether social justice and human rights should be foundational to AI governance from the outset or integrated as part of a broader governance process
The UN Pension Fund implemented a responsible digital identity solution for proof of life, converging AI, blockchain, and biometric facial recognition, and achieved the first ISO 42001 certificate on responsible use of AI within the UN system - Responsible AI implementation and certification (Dino Cataldo Dell'Accio)
Arg. 1Dino Cataldo Dell'Accio presents the UN Pension Fund's digital identity solution as a model for responsible and ethical AI implementation within the UN system. The solution replaced a 70-year-old paper-based process, converging multiple technologies, and the Fund became a pioneer by achieving the first ISO 42001 certificate on responsible use of AI.
He described the implementation in 2021 of a digital identity solution for Proof of Life, enabled with support from UNICC, which allowed 80,000 retirees in more than 192 countries to confirm they are still alive digitally, replacing a 70-year-old paper-based process . He noted that the application gave birth to the United Nations Digital Identity Solution, now joined by 16 UN entities . He stated that the Pension Fund demonstrated convergence of AI, blockchain, and biometric facial recognition , and became a pioneer by achieving the first ISO 42001 certificate on the responsible use of AI, demonstrating transparent implementation of AI systems .
UNICC provides shared digital infrastructure for the UN system, including a private cloud, shared digital identity systems, and treasury platforms, and is developing an AI factory, AI gateway, and AI studio to extend this shared approach to AI - Shared UN digital and AI infrastructure (Milena Grecuccio)
Arg. 1Milena Grecuccio argues that UNICC's role is to build and maintain the shared digital foundation for the UN system, ensuring that AI principles are translated into good practice through competent technical implementation. She outlines UNICC's plans to extend its shared infrastructure approach to AI through three conceptual pillars: an AI factory, AI gateway, and AI studio.
She described UNICC's shared digital infrastructure including a modern private cloud keeping data secure and protected by UN privileges and immunities, shared digital identity systems initiated by UNJSP, and treasury platforms under UNHCR leadership that move funds securely across borders . She outlined three pillars for extending this approach to AI: an AI factory for moving solutions safely from experimentation to deployment at scale, an AI gateway for governing transparently how AI is adopted including legal vetting, and an AI studio for building production-grade AI agents with guardrails embedded .
UNICRI's AI for Safe Children project provides over 100 AI tools to representatives from 129 countries to tackle online child sexual exploitation, resulting in saving over 100 children and arresting over 250 perpetrators - AI for child protection and law enforcement (Irakli Beridze)
Arg. 1Irakli Beridze presents UNICRI's AI for Safe Children project as a concrete example of AI delivering measurable impact in the area of child protection and law enforcement. The project provides AI tools to law enforcement officers worldwide to tackle online child sexual exploitation, with documented operational results.
He described the AI for Safe Children project, undertaken in partnership with the Ministry of Interior of the United Arab Emirates, through which UNICRI built an AI global hub providing over 100 AI tools to representatives of 129 countries to tackle the sexual exploitation and abuse of children online . He stated that through this project, over 3,500 law enforcement officers from 60 countries have been trained , and that operational undertakings following these trainings resulted in saving over 100 children, arresting over 250 perpetrators and sex offenders . He also mentioned a recent project with OpenAI Foundation on building curriculum on children safe online .
UNIDO's Global Alliance on AI for Industry Manufacturing (AIM Global) helps countries adopt industrial AI responsibly in the real economy, particularly in the Global South, through centres of excellence spanning multiple regions - Industrial AI for inclusive economic development (Jason Slater)
Arg. 1Jason Slater presents UNIDO's AIM Global initiative as a practical, partnership-based response to the challenge of bringing AI to the real economy, particularly in the Global South. He describes a network of centres of excellence that demonstrate, adapt, and apply industrial AI applications while building local skills and capacity, especially for SMEs.
He described the Global Alliance on AI for Industry Manufacturing (AIM Global), which brings together governments, industry, and academia to help countries adopt industrial AI responsibly in the real economy, especially in the Global South . He noted that AIM Global has built a partnership base of around 500 in over 70 countries, with centres of excellence spanning Africa, Asia, Europe, Latin America, and the Caribbean . He described the purpose of the centres as bringing technology to demonstrate, adapt, and apply industrial AI applications, build local skills and capacity particularly for SMEs, and provide a trusted space with clear guardrails .
UNIDO joined forces with ITU on Partner Connect and the Skills Coalition, connecting ITU's work on connectivity and skills with UNIDO's focus on industrial AI applications - Cross-agency partnership for industrial AI skills (Jason Slater)
Arg. 2Slater highlights UNIDO's partnership with ITU as an example of cross-agency collaboration that connects complementary areas of expertise. By joining forces on Partner Connect and the Skills Coalition, UNIDO and ITU are linking work on connectivity and skills with industrial AI applications to create greater impact.
He stated that UNIDO joined forces with ITU on Partner Connect and the Skills Coalition, connecting ITU's work on connectivity and skills across UNIDO's focus on industrial applications . He also noted that UNIDO launched a global call for solutions on the inclusive digital economy with UNCTAD, the UN Office for Digital Emerging Technologies, and the UN Development Coordination Office, with solutions feeding back into the centres of excellence .
The WSIS forum has brought together a diverse range of UN leaders to discuss leadership for inclusive, trusted, and sustainable digital transformation - Framing the panel discussion (Doreen Moderator)
Arg. 1In her role as moderator, Doreen frames the panel discussion around the theme of leadership for inclusive, trusted, and sustainable digital transformation, introducing each panellist and directing questions to them. She contextualises the discussion within the broader WSIS forum activities of the day, including ministerial roundtables and WSIS prize recognition.
She noted that the day had included a network of women leaders, an incredible ministerial roundtable, and recognition of WSIS prize winners , and that the panel would discuss leadership for inclusive, trusted, sustainable digital transformation . She introduced each panellist and directed specific questions to them, including asking Dr. Tedros to address youth and child protection in the age of AI .
Trust must extend beyond systems and data to encompass trust amongst UN agencies themselves, as this is how the UN system can succeed in avoiding duplication and competition - Trust as the foundation of UN cooperation (Doreen Moderator)
Arg. 2In her closing remarks to the first panel, Doreen synthesises the common themes she heard across panellists' interventions, identifying trust as a recurring and multidimensional concept. She argues that trust must not only apply to systems and data but must also characterise the relationships between UN agencies themselves.
She noted that there were many common references to trust across panellists' interventions, and that this trust must extend to trust amongst the UN agencies themselves . She explicitly stated that the system can succeed by ensuring they are not duplicating or competing, but working together to deliver better for member states and the multi-stakeholder community .
Digital transformation is having measurable impact across trade, postal services, climate prediction, health, transport standards, and women's empowerment, demonstrating the breadth of the UN system's work - Breadth of digital transformation impact (Doreen Moderator)
Arg. 3Doreen synthesises the contributions of the first panel by highlighting the wide range of sectors in which digital transformation is making a measurable difference, from trade and postal services to climate prediction, health, transport standards, and women's empowerment. This framing underscores the cross-cutting nature of the WSIS mandate.
She summarised that the panel had shown how digital transformation is impacting trade, the postal service, and how the post can be leveraged in connecting communities . She also noted the impact on climate prediction and early warnings, health and digital health foundations, transport standards and mobility, and empowering women .
The second panel focuses on implementation and delivery of concrete AI and digital-driven projects with measurable impact, moving from strategy to action - Framing the implementation panel (Moderator)
Arg. 1The moderator frames the second panel as a shift from the strategic vision discussed in the first panel to concrete implementation, asking each panellist to share examples of AI or digital-driven projects with measurable impact that can be replicated across the UN system and regions. This framing positions the second panel as the operational complement to the first.
She stated that the first panel had covered the policy side, and that the second panel would move to implementation aspects . She posed the same question to all panellists: to share examples of AI or digital-driven projects that have been implemented with measurable impact on the ground and that can be replicated across the UN system and across regions .
The UN General Assembly resolution asked the WSIS process to include human rights in all WSIS action lines, and the WSIS process is working with action line facilitators to ensure future roadmaps incorporate human rights - Human rights integration into WSIS action lines (Moderator)
Arg. 2The moderator highlights that a UN General Assembly resolution in December explicitly asked the WSIS process to integrate human rights across all WSIS action lines. She notes that work is underway with WSIS action line facilitators to ensure that future roadmaps reflect this requirement.
She stated that the UN General Assembly resolution in December asked the WSIS process to include human rights in all the WSIS action lines , and that the WSIS process is working with action line facilitators to ensure that future roadmaps include human rights while implementing them .
Mental health and well-being of children has been a major concern throughout the WSIS Forum, with multiple sessions dedicated to the topic, making UNICEF's child-centric AI benchmark work particularly relevant - Child well-being as a WSIS priority (Moderator)
Arg. 3The moderator contextualises UNICEF's presentation by noting that mental health and well-being of children was a significant theme throughout the WSIS Forum, with several dedicated sessions. She draws a direct link between this broader forum concern and UNICEF's work on child-centric AI benchmarks.
She noted that if one looked at the agenda of the WSIS Forum, mental health and well-being of children was a major part of the discussions throughout the week, with several sessions talking about how big a concern it had become . She described UNICEF's child-centric AI benchmark work as a great link to this broader forum theme .
UNESCO's work is so central to the WSIS vision that the World Summit on Information Society is often expanded to include 'Knowledge Societies' to reflect UNESCO's contribution - UNESCO and knowledge societies (Moderator)
Arg. 4The moderator acknowledges the breadth of UNESCO's contribution to the WSIS process by noting that the forum often expands the name of the World Summit on Information Society to include 'Knowledge Societies', reflecting the importance of UNESCO's work on knowledge, education, and culture within the WSIS framework.
She stated that within the WSIS process, the name World Summit on Information Society is often expanded to World Summit on Information and Knowledge Societies to include the work that UNESCO has been doing .
The WSIS process is the UN in action, with multiple UN agencies implementing work on the ground to have an impact on people all over the world - WSIS as demonstration of UN unity (Gitanjali)
Arg. 1In her closing remarks, Gitanjali frames the WSIS process as a concrete demonstration of the UN system working together in action, with agencies across the system implementing programmes that deliver real impact for people worldwide. She invites the assembled UN representatives for a group photograph as a symbol of this unity.
She stated that the WSIS process is really the UN in action, and that the audience had heard from several UN agencies about how they are implementing their work on the ground so that it can have an impact on people all over the world . She invited all UN agencies present for a group photograph to serve as a memory for the next WSIS Forum .
Session Knowledge Graph
Speakers · Topics · Arguments · Relationships
Dr. Tedros explicitly stated that 'the digital divide is a health divide' , warning that unless gaps in connectivity, capacity, governance, and financing are closed, digital transformation risks widening inequalities . Michelle Gyles-McDonnough introduced the concept of the 'capacity divide', noting that beyond the digital divide, countries now confront a capacity divide where access to technology alone is insufficient without skills, institutions, leadership, and policy capabilities . Tatiana Molcean echoed this, stressing that all countries must have the right skills and capacities to process and benefit from digital solutions . Pedro Manuel Moreno illustrated the gap concretely, noting that in most of the developing world, mobile phones mostly entertain rather than employ , and that when enabling conditions such as payment systems, logistics, legal frameworks, and skills are absent, people keep scrolling and income goes elsewhere . Sonja Leighton-Kone warned that the real risk is that the digital agenda moves fastest for those already connected, leaving the hardest to reach further behind . Doreen Bogdan-Martin framed the urgency by noting that 2.2 billion people remain offline .
Digital divide as health divide (Dr. Tedros Adhanom Ghebreyesus)
The capacity divide (Michelle Gyles-McDonnough)
Trustworthiness of data and national capacity (Tatiana Molcean)
Phone as economic tool vs. entertainment device (Pedro Manuel Moreno)
Digital inclusion for hardest-to-reach communities (Sonja Leighton-Kone)
With 2.2 billion people still offline, the digital divide remains a critical challenge requiring targeted action across the UN system - Scale of the digital divide (Doreen Bogdan-Martin)
Celeste Saulo articulated the foundational principle that 'AI is only as good as the data behind it' , demonstrating through WMO's experience that AI models for drought detection cannot succeed without decades of internationally shared observational data . She described WMO's 2021 unified data policy, unanimously adopted by members, committing to exchange Earth system data freely and without restriction , and the WIS2 platform as the open backbone of all forecasting including AI . Tatiana Molcean joined this point, calling the trustworthiness of data essential when talking about digital transformation . Milena Grecuccio argued that the tech adopted and developed by the UN must embed technical, architectural governance and human rights principles by design, consistently across the system in a transparent and trusted manner . Sonja Leighton-Kone stated that robust data frameworks and digital infrastructure are absolute prerequisites for safe AI deployment .
Data foundations for trustworthy AI (Celeste Saulo)
Trustworthiness of data and national capacity (Tatiana Molcean)
Shared UN digital and AI infrastructure (Milena Grecuccio)
Multi-tenant shared operational systems (Sonja Leighton-Kone)
Magdalena Sepulveda stated that at UNRISD, when many discussions on AI speak about technology, they speak about people, asking who benefits, who is left behind, and what policies are needed , and concluded that if AI is going to contribute to sustainable development, social justice cannot be an afterthought but must be included from the very beginning . Dr. Tedros argued that digital transformation is not about technology alone but about helping countries build stronger health systems and enabling people everywhere to access better health services , and that AI must be governed responsibly with people at the centre . Kaan Cetinturk insisted that scale and impact must be measured not only by numbers but also through the lens of equity, and that digital transformation should empower the most vulnerable . Michelle Gyles-McDonnough emphasised that digital transformation succeeds only when countries have the capacity not only to access technology but also to understand it, govern it, and apply it to their development priorities . Miguel Panadero concluded that the last mile of digital transformation is not technology but people .
Social justice in AI governance (Magdalena Sepulveda)
Digital cooperation for health (Dr. Tedros Adhanom Ghebreyesus)
Equity lens for digital transformation (Kaan Cetinturk)
The capacity divide (Michelle Gyles-McDonnough)
Building digital capabilities within the UN system (Miguel Panadero)
Michelle Gyles-McDonnough argued that closing the capacity divide must be a strategic priority and not the add-on that is usually seen , with UNITAR reaching almost 600,000 learners across 105 countries each year . Mariya Gabriel described UNESCO training more than 35,000 civil servants through the UNESCO-Oxford Programme on AI and Digital Transformation in Government , and operating the Spark AI Alliance across more than 100 countries . Miguel Panadero described the UN System Staff College's role in turning digital ambition into real institutional capability , with the Blue Line Learning Marketplace connecting tens of thousands of UN colleagues worldwide , and UNSDG Learn reaching more than 1 million users worldwide . Beate Andrees highlighted ILO's International Training Centre in Turin, which has set up capacity-building programmes for labour ministries and workers' and employers' organisations . Robert Opp noted UNDP conducts capacity building in over 80 countries with civil servants, civil society, and community groups . Tatiana Molcean stressed the importance of supporting all countries to ensure they have the right skills and capacities .
Capacity building for policymakers (Michelle Gyles-McDonnough)
Civil servant AI training at scale (Mariya Gabriel)
Building digital capabilities within the UN system (Miguel Panadero)
Labour market capacity building and governance (Beate Andrees)
Broad-based digital capacity building (Robert Opp)
Trustworthiness of data and national capacity (Tatiana Molcean)
Doreen Bogdan-Martin stressed that the UN must ensure it is not duplicating or competing but working together to deliver better for member states and the multi-stakeholder community , identifying trust amongst UN agencies themselves as fundamental . Tomas Lamanauskas described UNGIS as bringing together 31 UN entities and 17 observers, making it nearly 50 entities working together , and noted that WSIS involved over 50,000 stakeholders from 160 countries . Pedro Manuel Moreno highlighted that as chair of UNGIS, UNCTAD tabled a joint implementation roadmap to align WSIS and the Global Digital Compact . Milena Grecuccio described UNICC as building the shared digital foundation for the UN system, with plans to extend this approach to AI through an AI factory, AI gateway, and AI studio . Sonja Leighton-Kone explained that instead of 40 separate agencies building 40 separate tools, UNOPS designs systems with multi-tenant architecture available to sister UN agencies . Dino Cataldo Dell'Accio noted that the UN Digital Identity Solution, born from the Pension Fund's proof-of-life application, is now joined by 16 UN entities . Jason Slater highlighted UNIDO joining forces with ITU on Partner Connect and the Skills Coalition .
The UN system must avoid duplication and competition, working together through UNGIS and shared platforms to deliver better for member states and the multi-stakeholder community - UN system cooperation imperative (Doreen Bogdan-Martin)
WSIS process is a uniquely collaborative UN effort involving over 50,000 stakeholders from 160 countries, with UNGIS coordinating system-wide delivery - UN system coordination (Tomas Lamanauskas)
UNCTAD, as chair of UNGIS, tabled a joint implementation roadmap to align WSIS and the Global Digital Compact, demonstrating concrete system-wide coordination - Aligning WSIS and Global Digital Compact (Pedro Manuel Moreno)
Shared UN digital and AI infrastructure (Milena Grecuccio)
Multi-tenant shared operational systems (Sonja Leighton-Kone)
Responsible AI implementation and certification (Dino Cataldo Dell'Accio)
Cross-agency partnership for industrial AI skills (Jason Slater)
Dr. Tedros stated that AI must be safe, ethical, evidence-based, and equitable, and must be governed responsibly with people at the centre , with WHO, ITU, and WIPO helping countries adopt AI with confidence through guidance, standards, and governance frameworks . Magdalena Sepulveda insisted that social justice cannot be an afterthought in AI governance but must be included from the very beginning . Mariya Gabriel described UNESCO's support for member states in implementing the Recommendation on the Ethics of AI, adopted by all 193 UNESCO member states, including AI readiness assessments, policy development, and guidance on bias, transparency, and accountability . Anwar Mahfoudh reframed human rights as not an obstacle to AI but the operating system that AI should be built upon safely , and that when built responsibly, AI can strengthen protection, accountability, and scale trusted decision-making across the UN . Milena Grecuccio argued that the tech adopted and developed by the UN must embed technical, architectural governance and human rights principles by design . Dino Cataldo Dell'Accio described the Pension Fund becoming a pioneer by achieving the first ISO 42001 certificate on the responsible use of AI . Celeste Saulo concluded that trustworthy AI does not begin with algorithms but with open data, standards, open source software, and the capacity to use them .
AI must be governed responsibly with people at the centre; WHO, ITU, and WIPO are helping countries adopt AI with confidence through guidance, standards, and governance frameworks - Responsible AI governance frameworks (Dr. Tedros Adhanom Ghebreyesus)
Social justice in AI governance (Magdalena Sepulveda)
Ethics of AI implementation (Mariya Gabriel)
Human rights as the foundation for AI (Anwar Mahfoudh)
Shared UN digital and AI infrastructure (Milena Grecuccio)
Responsible AI implementation and certification (Dino Cataldo Dell'Accio)
Data foundations for trustworthy AI (Celeste Saulo)
Doreen Bogdan-Martin stated that it is the collective responsibility of the UN family to help carry the WSIS vision and the renewed mandate of the WSIS to 2035 , and that the past 20 years has proven that multi-stakeholder cooperation works . Dr. Tedros affirmed that the renewal of the WSIS mandate sends an important message that digital cooperation remains essential to achieving sustainable development . Tomas Lamanauskas described the WSIS process as a genuinely unified UN process over the years, bringing all the UN system around the goals , and noted that WSIS Plus 20 assigned UNGIS additional work including on financing for digital development . Gitanjali concluded that the WSIS process is really the UN in action, with several UN agencies implementing their work on the ground to have an impact on people all over the world .
UN family leads WSIS action lines and must carry forward the renewed WSIS mandate to 2035 through multi-stakeholder cooperation - Opening remarks on UN unity (Doreen Bogdan-Martin)
Digital cooperation for health (Dr. Tedros Adhanom Ghebreyesus)
WSIS process is a uniquely collaborative UN effort involving over 50,000 stakeholders from 160 countries, with UNGIS coordinating system-wide delivery - UN system coordination (Tomas Lamanauskas)
WSIS as demonstration of UN unity (Gitanjali)
These three speakers share the view that AI and digital tools must be built upon robust, real-world data and physical infrastructure to deliver meaningful outcomes for vulnerable populations. Celeste Saulo demonstrated that AI models for drought detection cannot succeed without decades of observational data contributed by countries worldwide , and that WMO has spent decades building the foundations that trustworthy AI requires . Kaan Cetinturk described UNICEF's Ahead of the Storm initiative, which uses AI to convert complex climate hazard forecasts and contextual vulnerability data into timely, decision-ready operational data for countries to prepare for emergencies affecting 1.1 billion children . Masahiko Metoki highlighted that post offices, with around 600,000 locations mostly in rural areas, provide the physical infrastructure through which digital services can reach those who may not easily access the internet, smartphones, or digital skills , while noting that around 100,000 post offices still lack meaningful internet connection . These four speakers share a focus on building human and institutional capacity as the critical enabler of digital transformation, each targeting different but complementary audiences. Michelle Gyles-McDonnough focused on policymakers, diplomats, and public servants through UNITAR's five areas of action, reaching almost 600,000 learners across 105 countries . Mariya Gabriel highlighted UNESCO's training of more than 35,000 civil servants through the UNESCO-Oxford Programme and the Spark AI Alliance operating in more than 100 countries . Miguel Panadero described the UN System Staff College's role in helping UN personnel build digital capabilities through learning programmes, shared infrastructure, and responsible AI tools, with UNSDG Learn reaching more than 1 million users worldwide . Beate Andrees outlined ILO's capacity-building programmes for labour ministries and workers' and employers' organisations through the International Training Centre in Turin , alongside the adoption of the first international treaty on decent work in the platform economy . These three speakers share the view that digital inclusion requires not just connectivity but a full ecosystem of enabling conditions, and that specific attention must be paid to the hardest-to-reach communities. Pedro Manuel Moreno argued that whether a phone is used for economic activity depends on things the user never sees — a payment system that works, a parcel service that delivers, a law that protects the buyer, and the skills to run a business . Masahiko Metoki highlighted that post offices can provide digital services cost-effectively to people who may not easily access the internet, smartphones, or digital skills , and that UPU will continue to connect communities through the post office in partnership with other UN organisations . Sonja Leighton-Kone stated that true inclusion requires foundational public infrastructure, institutional capability, and data governance that makes connectivity useful once it is there , and that UNOPS brings digital transformation and emergency connectivity directly to post-conflict LDCs and small island development states . These four speakers share the view that human rights, social justice, and ethical principles must be foundational to AI governance rather than supplementary considerations. Anwar Mahfoudh argued that human rights are not an obstacle to AI but the operating system that AI should be built upon safely , and that the Human Rights Data Exchange, powered by generative AI, makes human rights information accessible, connected, and interoperable . Magdalena Sepulveda insisted that the measure of success of AI cannot be determined solely by what AI can do but must be measured by how AI contributes to societies , and that social justice cannot be an afterthought . Dr. Tedros stated that AI must be safe, ethical, evidence-based, and equitable, and must be governed responsibly with people at the centre . Mariya Gabriel described UNESCO's support for implementing the Recommendation on the Ethics of AI, adopted by all 193 member states, including guidance on bias, transparency, and accountability . These speakers share a commitment to building shared, interoperable UN digital infrastructure and governance mechanisms that avoid duplication and create system-wide efficiencies. Tomas Lamanauskas described UNGIS as bringing together nearly 50 entities to ensure coherent delivery , and noted it has been recognised by member states as an important platform . Pedro Manuel Moreno highlighted the joint implementation roadmap to align WSIS and the Global Digital Compact as a concrete example of system-wide coordination . Milena Grecuccio described UNICC as building the shared digital foundation for the UN system, with plans to extend this to AI through an AI factory, AI gateway, and AI studio . Sonja Leighton-Kone explained that UNOPS designs multi-tenant architecture systems available to sister UN agencies rather than each building separate tools . Dino Cataldo Dell'Accio noted that the UN Digital Identity Solution, born from the Pension Fund's initiative, is now joined by 16 UN entities, demonstrating collaboration, interoperability, and shared efficiencies . These speakers share the view that digital and AI transformation must translate into tangible economic benefits, particularly for developing countries and vulnerable workers, and that the next phase of the WSIS mandate will be judged by economic outcomes. Jason Slater argued that responsible AI must reach factories, firms, and workers, not only laboratories and platforms , with UNIDO's AIM Global initiative helping countries adopt industrial AI responsibly in the real economy, especially in the Global South . Robert Opp described UNDP hard-coding digital and AI transformation into its 2026-2029 strategic plan as a cross-cutting accelerator . Beate Andrees highlighted ILO's adoption of the first-ever international treaty on decent work in the platform economy and collaborative research on AI's macroeconomic impact . Pedro Manuel Moreno concluded that the first 20 years of the summit put much of the world online, and the next 10 will be judged by what being online is worth — a sale, a job, an income .
In a forum dominated by discussion of cutting-edge AI and digital innovation, there was unexpected consensus among speakers from very different organisations that legacy and physical infrastructure remains indispensable for digital inclusion. Masahiko Metoki acknowledged that when people hear the word 'postal', they might feel it is old-fashioned and apart from digital, IT, and AI , yet argued that the WSIS Plus 20 document explicitly recognises post offices as inclusive access points for delivering digital services in rural areas , with around 600,000 post offices worldwide mostly in rural areas . Celeste Saulo similarly noted that WMO has spent decades building data-sharing foundations through what were once ageing systems, and that WIS2 is the open successor serving as the backbone of all forecasting including AI . Sonja Leighton-Kone argued that true inclusion requires foundational public infrastructure, institutional capability, and data governance that makes connectivity useful once it is there , and that UNOPS brings digital transformation directly to post-conflict LDCs and small island development states . The consensus that old infrastructure must be upgraded rather than bypassed was unexpected given the forward-looking AI focus of the forum.
In a year when, as Celeste Saulo noted, 'cooperation feels harder everywhere' , there was unexpected consensus that technical cooperation through open standards can succeed even when political cooperation is strained. Saulo described how 11 countries including the United States, China, Brazil, Canada, France, Germany, Japan, Morocco, Saudi Arabia, South Korea, and the United Kingdom operate the global backbone of WIS2, noting that 'whatever may divide these countries elsewhere, here they build together - pure cooperation through technology for the common good' . She explained that rather than compelling nations into a single system, WMO agreed on a common language of open standards that allow each country to stay independent while operating as a global network . Tomas Lamanauskas described the WSIS process as a uniquely truly one UN process, with over 50,000 stakeholders from 160 countries involved . Milena Grecuccio described UNICC as a consortium where every UN organisation has a voice and the ability to steer the technology and services delivered, in a collective for good direction . This consensus that technical standards and shared infrastructure can bridge geopolitical divides was particularly striking given the broader context of international tensions.
There was unexpected consensus, particularly from a human rights body, that human rights frameworks should be understood as the foundational operating system for AI rather than as regulatory obstacles. Anwar Mahfoudh explicitly reframed the relationship, stating that 'human rights for us, we feel that they are not an obstacle to AI. They are the operating system that AI should be built upon safely' , and that when built responsibly, AI can strengthen protection, accountability, and scale trusted decision-making across the UN . This framing was echoed by Dr. Tedros, who argued that AI must be governed responsibly with people at the centre , and by Magdalena Sepulveda, who insisted that the measure of success of AI must be measured by how AI contributes to societies . Milena Grecuccio argued that the tech adopted and developed by the UN must embed technical, architectural governance and human rights principles by design . The consensus that human rights are a positive foundation for AI rather than a constraint was notable in a technical forum context.
It was unexpected that operational and administrative bodies - the UN Pension Fund, UNICC, and UNOPS - emerged as pioneers of responsible AI adoption and shared digital infrastructure, rather than the more programmatic or policy-focused agencies. Dino Cataldo Dell'Accio described the UN Pension Fund becoming a pioneer by achieving the first ISO 42001 certificate on the responsible use of AI within the UN system , and the Proof of Life digital identity solution giving birth to the United Nations Digital Identity Solution now joined by 16 UN entities . Milena Grecuccio described UNICC as having built and continuing to build the shared digital foundation for the UN system over 55 years , and now developing an AI factory, AI gateway, and AI studio to extend this approach to AI . Sonja Leighton-Kone described UNOPS designing multi-tenant architecture systems available to sister UN agencies, enabling shared interoperable approaches rather than 40 separate agencies building 40 separate tools . The consensus that back-office and operational bodies are driving system-wide digital innovation was unexpected and significant.
The discussion revealed a remarkably high level of consensus across all participating UN agencies on the fundamental principles and priorities for digital transformation and AI governance. Key areas of agreement included: the multidimensional nature of the digital divide extending beyond connectivity to capacity, skills, and governance; the necessity of placing people and social justice at the centre of AI governance; the importance of trustworthy, open data as the foundation for AI; the strategic priority of capacity building for policymakers and institutions; the imperative of UN system-wide cooperation through shared infrastructure and UNGIS to avoid duplication; and the need for responsible, ethical AI governance with human rights embedded from the outset. The WSIS mandate renewal to 2035 was universally affirmed as the appropriate framework for advancing these goals. Concrete examples of cross-agency collaboration - such as the WHO-ITU-WIPO joint publication , the UN Digital Identity Solution joining 16 entities , the E-Trade for All initiative with 35 organisations , and the joint WSIS-GDC implementation roadmap - demonstrated that this consensus is translating into operational reality. The second panel reinforced these themes by showcasing measurable, replicable AI and digital projects across health, climate, trade, child protection, human rights, and industrial development.
Doreen Bogdan-Martin frames the central challenge as one of connectivity, citing 2.2 billion people still offline , implying that access remains the primary unresolved issue. Pedro Manuel Moreno, however, argues that the more pressing challenge is what being online is actually worth economically, noting that in most of the developing world, phones mostly entertain rather than employ , and that the first 20 years put much of the world online while the next 10 will be judged by what being online is worth . Sonja Leighton-Kone partially bridges these positions by arguing that true inclusion requires not just connectivity but foundational public infrastructure, institutional capability, and data governance , suggesting that connectivity alone is insufficient even when achieved. These represent meaningfully different diagnoses of where the primary problem lies.
With 2.2 billion people still offline, our work is far from over - Scale of the digital divide (Doreen Bogdan-Martin)
UNCTAD's E-Trade for All initiative, now with 35 organisations, provides developing countries a single entry point into the UN system's digital support, and the next decade will be judged by what being online is worth economically - Digital trade and economic inclusion (Pedro Manuel Moreno)
UNOPS brings digital transformation and emergency connectivity directly to post-conflict least developed countries and small island developing states, focusing on building capability rather than dependency - Digital inclusion for hardest-to-reach communities (Sonja Leighton-Kone)
Celeste Saulo explicitly cautions against treating AI as a primary driver, stating that AI is only as good as the data behind it and that AI does not create resilience by itself but merely enhances it , emphasising that trustworthy AI begins with open data, standards, and open source software rather than algorithms . By contrast, Robert Opp positions AI and digital transformation as a hard-coded strategic accelerator across all of UNDP's work , treating AI as a central driver rather than a dependent variable. Kaan Cetinturk further complicates this by arguing that the measure of AI's success must go beyond scale and numbers to include equity , implicitly challenging approaches that treat AI deployment volume as a primary indicator of progress.
AI is only as good as the data behind it; WMO's decades of open Earth system data sharing through its unified data policy and WIS2 platform provide the trusted foundation that AI in meteorology requires - Data foundations for trustworthy AI (Celeste Saulo)
UNDP has hard-coded digital and AI transformation into its 2026–2029 strategic plan as a cross-cutting accelerator, supporting both horizontal digital governance and vertical thematic programme integration - AI and digital as strategic accelerators for development (Robert Opp)
UNICEF's digital transformation initiatives must be measured through the lens of equity, ensuring they empower the most vulnerable, particularly children, and bridge rather than widen the digital divide - Equity lens for digital transformation (Kaan Cetinturk)
Doreen Bogdan-Martin frames the overarching challenge primarily in terms of the digital divide, citing 2.2 billion people still offline as evidence that the work is far from over . Michelle Gyles-McDonnough, however, introduces a distinct and arguably more urgent challenge: the capacity divide, referencing the Independent International Scientific Panel on AI report to argue that beyond the digital divide, countries are now confronting a capacity divide where access to technology alone is not enough . She argues that countries need skills, institutions, leadership, and policy capabilities to harness digital technologies responsibly . Tatiana Molcean reinforces this by stressing that all countries must have the right skills and capacities to process and benefit from digital solutions , and frames this as one way to solve inequality . These positions represent a meaningful tension between prioritising connectivity versus prioritising capability.
With 2.2 billion people still offline, our work is far from over - Scale of the digital divide (Doreen Bogdan-Martin)
Beyond the digital divide, there is now a capacity divide; access to technology alone is insufficient without the skills, institutions, leadership, and policy capabilities to harness it responsibly - The capacity divide (Michelle Gyles-McDonnough)
Trustworthiness of data is essential for digital transformation, and all countries must have the right skills and capacities to process and benefit from digital solutions - Data trustworthiness and national capacity (Tatiana Molcean)
Magdalena Sepulveda makes the strongest claim, arguing that social justice cannot be an afterthought in AI governance but must be included from the very beginning , and that the measure of AI's success cannot be determined solely by what AI can do but must be measured by how it contributes to societies . Anwar Mahfoudh similarly argues that human rights are not an obstacle to AI but the operating system upon which AI should be built . Mariya Gabriel, while supportive of ethics, frames UNESCO's approach more procedurally through readiness assessments, policy development, and guidance on bias and transparency , which implies a process of integration rather than a foundational prerequisite. This represents a subtle but meaningful difference between treating social justice and human rights as constitutive of AI governance versus treating them as important considerations to be incorporated through established processes.
UNRISD focuses on the social dimensions of AI, asking who benefits, who is left behind, and what policies are needed; social justice must be included from the very beginning of AI governance, not as an afterthought - Social justice in AI governance (Magdalena Sepulveda)
Human rights are not an obstacle to AI but the operating system upon which AI should be built; OHCHR launched the Human Rights Data Exchange to make scattered human rights data accessible and interoperable across the UN ecosystem - Human rights as the foundation for AI (Anwar Mahfoudh)
UNESCO supports member states in implementing its Recommendation on the Ethics of AI, adopted by all 193 member states, including AI readiness assessments, policy development, and guidance on bias, transparency, and accountability - Ethics of AI implementation (Mariya Gabriel)
In a forum where speakers frequently cited large numbers to demonstrate impact - such as UNICC's tens of thousands of connected UN colleagues , UNITAR's 600,000 learners , and the Pension Fund's 80,000 retirees in 192 countries - Kaan Cetinturk explicitly challenged this approach, arguing that scale and impact must be measured not only by number of people, number of countries, and numbers generally, but also through the lens of equity . This was unexpected in a context where other speakers, including Milena Grecuccio who described UNICC's shared infrastructure reaching the entire UN system , and Dino Cataldo Dell'Accio who highlighted the Pension Fund's achievement of the first ISO 42001 certificate , were implicitly using scale and technical achievement as primary indicators of success. Cetinturk's challenge to the numbers-based framing was a subtle but notable departure from the dominant rhetorical approach of the panel.
In a forum where the general tone was enthusiastic about AI's transformative potential, Celeste Saulo offered an unexpectedly cautionary framing, explicitly stating that AI is only as good as the data behind it and that AI does not create resilience by itself but merely enhances it . She further argued that trustworthy AI does not begin with algorithms but with open data, standards, open source software, and the capacity to use them . This contrasted sharply with Robert Opp's framing of AI and digital transformation as a hard-coded strategic accelerator across all of UNDP's work , and Jason Slater's presentation of industrial AI as a clear driver of inclusive and sustainable industrial development . The unexpected element is that Saulo's position - that AI is derivative of pre-existing cooperation and data infrastructure rather than a primary driver - implicitly challenges the AI-centric framing adopted by most other speakers, though this tension was not directly addressed in the discussion.
An unexpected tension emerged between speakers focused on the UN system's internal digital transformation - Panadero on building digital capabilities within UN personnel , Grecuccio on shared UN digital infrastructure , and Dell'Accio on the Pension Fund's internal AI implementation - and the broader panel framing led by Doreen Bogdan-Martin, which emphasised the UN system's role in delivering digital transformation for member states and the world . While not a direct confrontation, the second panel's focus on internal UN systems and processes sat in some tension with the first panel's outward-facing development mandate. Bogdan-Martin's call to avoid duplication and competition implicitly acknowledged that the proliferation of separate internal digital systems across agencies - which Sonja Leighton-Kone explicitly noted as a problem - represents a real challenge that the internal-focused presentations did not fully resolve.
The discussion was characterised by a high degree of surface-level consensus around shared goals - digital inclusion, responsible AI, multi-stakeholder cooperation, and the WSIS mandate - but contained meaningful underlying tensions around four key issues: (1) whether connectivity or economic utility of connectivity is the primary challenge ; (2) whether AI is a primary driver or merely an enhancer dependent on pre-existing data and cooperation foundations ; (3) whether the digital divide or the capacity divide is the more urgent priority ; and (4) whether social justice and human rights should be constitutive of AI governance from the outset or integrated through established processes . Additionally, a subtle tension existed between the UN system's internal digital transformation agenda and its external development mandate .
All four speakers agree that digital transformation risks widening rather than reducing inequalities if foundational gaps are not addressed. Dr. Tedros states explicitly that unless gaps in connectivity, capacity, governance, and financing are closed, digital transformation risks widening inequalities . Celeste Saulo agrees that AI enhances rather than creates resilience and depends on foundational data infrastructure . Tatiana Molcean echoes the trustworthiness of data point and the need for skills and capacities . Michelle Gyles-McDonnough frames this as the capacity divide . However, they disagree on which gap is most critical: Dr. Tedros emphasises connectivity and financing , Saulo emphasises data quality and open standards , Molcean emphasises skills and data trustworthiness , and Gyles-McDonnough emphasises institutional and policy capabilities .
The digital divide is also a health divide; unless gaps in connectivity, capacity, governance, and financing are closed, digital transformation risks widening inequalities - Digital divide as health divide (Dr. Tedros Adhanom Ghebreyesus) AI is only as good as the data behind it; WMO's decades of open Earth system data sharing through its unified data policy and WIS2 platform provide the trusted foundation that AI in meteorology requires - Data foundations for trustworthy AI (Celeste Saulo) Trustworthiness of data is essential for digital transformation, and all countries must have the right skills and capacities to process and benefit from digital solutions - Data trustworthiness and national capacity (Tatiana Molcean) Beyond the digital divide, there is now a capacity divide; access to technology alone is insufficient without the skills, institutions, leadership, and policy capabilities to harness it responsibly - The capacity divide (Michelle Gyles-McDonnough)
All four speakers agree that digital inclusion must prioritise the most vulnerable and hardest-to-reach populations. Metoki focuses on rural communities, elderly people, and women through post offices , Moreno focuses on developing-world users who need economic enablers beyond connectivity , Leighton-Kone focuses on post-conflict LDCs and small island developing states , and Cetinturk focuses on children as the most vulnerable . However, they disagree on the mechanism: Metoki advocates leveraging existing physical postal infrastructure , Moreno argues for building the full ecosystem of payment systems, logistics, legal frameworks, and skills , Leighton-Kone emphasises building local capability rather than dependency , and Cetinturk emphasises measuring impact through an equity lens .
Post offices, with around 600,000 locations worldwide mostly in rural areas, can provide digital services cost-effectively to rural communities, elderly people, and women through initiatives like Post for Women - Post offices bridging rural digital divide (Masahiko Metoki) In most of the developing world, mobile phones are used predominantly for entertainment rather than economic activity; closing this gap requires functional payment systems, logistics, legal frameworks, and skills - Phone as economic tool vs. entertainment device (Pedro Manuel Moreno) UNOPS brings digital transformation and emergency connectivity directly to post-conflict least developed countries and small island developing states, focusing on building capability rather than dependency - Digital inclusion for hardest-to-reach communities (Sonja Leighton-Kone) UNICEF's digital transformation initiatives must be measured through the lens of equity, ensuring they empower the most vulnerable, particularly children, and bridge rather than widen the digital divide - Equity lens for digital transformation (Kaan Cetinturk)
All four speakers agree that AI must be governed responsibly and with people at the centre. Dr. Tedros states that AI must be safe, ethical, evidence-based, and equitable, and must be governed responsibly with people at the centre . Sepulveda argues that the measure of AI's success must be how it contributes to societies . Mahfoudh frames human rights as the operating system for AI . Gabriel references the UNESCO Recommendation on the Ethics of AI adopted by 193 member states . However, they differ on the mechanism: Dr. Tedros emphasises guidance, standards, and governance frameworks developed collaboratively , Sepulveda emphasises generating evidence about inequality and broadening participation in governance discussions , Mahfoudh emphasises making human rights data accessible and interoperable , and Gabriel emphasises readiness assessments and policy development tools .
AI must be governed responsibly with people at the centre; WHO, ITU, and WIPO are helping countries adopt AI with confidence through guidance, standards, and governance frameworks - Responsible AI governance frameworks (Dr. Tedros Adhanom Ghebreyesus) UNRISD focuses on the social dimensions of AI, asking who benefits, who is left behind, and what policies are needed; social justice must be included from the very beginning of AI governance, not as an afterthought - Social justice in AI governance (Magdalena Sepulveda) Human rights are not an obstacle to AI but the operating system upon which AI should be built; OHCHR launched the Human Rights Data Exchange to make scattered human rights data accessible and interoperable across the UN ecosystem - Human rights as the foundation for AI (Anwar Mahfoudh) UNESCO supports member states in implementing its Recommendation on the Ethics of AI, adopted by all 193 member states, including AI readiness assessments, policy development, and guidance on bias, transparency, and accountability - Ethics of AI implementation (Mariya Gabriel)
All three speakers agree that UNGIS is the critical coordination mechanism for the UN system's digital work and that system-wide cooperation must be strengthened. Lamanauskas describes UNGIS as bringing together 31 UN entities and 17 observers and notes it has been given additional responsibilities under WSIS Plus 20 . Moreno highlights UNCTAD's role as UNGIS chair in tabling a joint implementation roadmap to align WSIS and the Global Digital Compact . Bogdan-Martin stresses the need to avoid duplication and competition . However, they differ in emphasis: Lamanauskas frames UNGIS as an example of broader UN reform , Moreno focuses on the concrete output of the joint roadmap , and Bogdan-Martin emphasises the trust dimension of inter-agency cooperation .
UNGIS brings together 31 UN entities and 17 observers to coordinate system-wide digital delivery, and has been tasked with additional work under WSIS Plus 20, including on financing for digital development - UNGIS as coordination mechanism (Tomas Lamanauskas) UNCTAD, as chair of UNGIS, tabled a joint implementation roadmap to align WSIS and the Global Digital Compact, demonstrating concrete system-wide coordination - Aligning WSIS and Global Digital Compact (Pedro Manuel Moreno) The UN system must avoid duplication and competition, working together through UNGIS and shared platforms to deliver better for member states and the multi-stakeholder community - UN system cooperation imperative (Doreen Bogdan-Martin)
All four speakers agree that capacity building is essential to digital transformation and that their respective organisations are contributing to this goal. However, they target different audiences and use different mechanisms: Gyles-McDonnough focuses on policymakers, diplomats, and public servants through executive education and AI governance training , reaching almost 600,000 learners across 105 countries ; Panadero focuses on UN personnel themselves through shared learning infrastructure and responsible AI tools , with UNSDG Learn reaching over 1 million users ; Andrees focuses on labour ministries and workers' and employers' organisations through the International Training Centre in Turin , alongside governance through international standards ; and Gabriel focuses on civil servants through the UNESCO-Oxford Programme training over 35,000 civil servants and the Spark AI Alliance across more than 100 countries . The partial disagreement lies in whether capacity building should prioritise UN internal staff, national civil servants, labour constituencies, or broader public administration.
UNITAR strengthens the capabilities of policymakers, diplomats, and public servants through executive education, AI governance training, and tailored advisory support, reaching almost 600,000 learners across 105 countries annually - Capacity building for policymakers (Michelle Gyles-McDonnough) The UN System Staff College helps UN personnel build digital capabilities through learning programmes, shared digital learning infrastructure such as the Blue Line Learning Marketplace, and responsible AI tools including a De-Biasing Coach - Building digital capabilities within the UN system (Miguel Panadero) ILO conducts collaborative research on AI's impact on occupations and labour markets, runs capacity-building programmes for labour ministries and workers' organisations, and has adopted the first international treaty on decent work in the platform economy - Labour market capacity building and governance (Beate Andrees) UNESCO is training over 35,000 civil servants through the UNESCO-Oxford Programme on AI and Digital Transformation in Government, and operates the Spark AI Alliance across more than 100 countries - Civil servant AI training at scale (Mariya Gabriel)
- The UN family, coordinated through UNGIS (UN Group on the Information Society), plays a central role in advancing the renewed WSIS mandate to 2035, with over 31 entities and 17 observers working collectively to deliver digital transformation for sustainable development.
- With 2.2 billion people still offline, the digital divide remains a critical and urgent challenge; it is simultaneously a health divide, an economic divide, and a capacity divide, requiring coordinated action across all UN agencies.
- AI presents unprecedented opportunities across sectors including health, climate, child protection, industrial development, and human rights, but must be safe, ethical, evidence-based, equitable, and governed responsibly with people at the centre.
- Trustworthy AI depends fundamentally on open, high-quality data; WMO's unified data policy and WIS2 platform exemplify how treating data as a global public good provides the foundation for reliable AI applications.
- Beyond connectivity, there is a growing 'capacity divide' — countries need skills, institutions, leadership, and policy capabilities to harness digital technologies responsibly, making capacity building a strategic priority rather than an add-on.
- Post offices, with approximately 600,000 locations worldwide mostly in rural areas, represent an underutilised but powerful channel for delivering digital services to rural communities, elderly people, and women, though an estimated 100,000 still lack meaningful internet connectivity.
- The UN system must avoid duplication and competition by leveraging shared infrastructure, interoperable platforms, and multi-tenant architectures — as demonstrated by UNICC's shared digital foundation, UNOPS's multi-tenant systems, and the UN Pension Fund's digital identity solution now adopted by 16 UN entities.
- Social justice and human rights must be embedded from the very beginning of AI governance, not treated as afterthoughts; human rights are described as the 'operating system' upon which AI should be built.
- The first 20 years of WSIS put much of the world online; the next decade (to 2035) will be judged by what being online is worth economically and socially — a job, a sale, an income, better health, or greater inclusion.
- Multi-stakeholder cooperation, as demonstrated through the WSIS process involving over 50,000 stakeholders from 160 countries, has proven effective over 20 years and remains the model for advancing inclusive digital transformation.
- The last mile of digital transformation is not technology but people; individual and collective leadership, skills, and institutional capability are the ultimate determinants of whether digital transformation delivers on its promises.
- UNCTAD tabled a joint implementation roadmap to align WSIS and the Global Digital Compact, representing a concrete step towards coherent system-wide digital governance.
“The digital divide is a health divide. Unless we close gaps in connectivity, capacity, governance and financing, digital transformation risks widening inequalities rather than reducing them.”
“AI is only as good as the data behind it. AI does not create resilience by itself. It enhances it.”
“The phone that you have in your pockets can either entertain you or can employ you. In most of the developing world, it mostly entertains.”
“Beyond the digital divide, we're now confronting the capacity divide. Access to technology alone is not enough. Countries need the skills, institutions, leadership, and policy capabilities to harness digital technologies responsibly and effectively.”
“Social justice cannot be an afterthought. It has to be included from the very beginning. The measure of success of AI cannot be determined solely by what AI can do, but has to be measured in regard to how AI is contributing to our societies.”
“The last mile of digital transformation is not technology. It's people.”
“Human rights are not an obstacle to AI. They are the operating system that AI should be built upon safely.”
“Only 35% of SDG targets are on track or making moderate progress. 18% of them are moving backwards. The question is, a lot of people think the only hope is technology — how we really leverage technology to leapfrog in different areas.”
How can the UN system ensure that digital transformation serves as an enabler for the UN reform process (UN80) without duplication or competition between agencies?
This question is central to the entire discussion and was raised at the outset. With multiple UN agencies working on overlapping digital agendas, understanding how to coordinate effectively is critical to maximising impact and avoiding waste of resources.
How can we leverage digital technologies to better serve member states and deliver on UN mandates collectively?
This broad strategic question underpins the panel discussion and was left partially answered. It requires further exploration of concrete mechanisms for system-wide coordination, particularly as the WSIS mandate extends to 2035.
With 2.2 billion people still offline, what specific strategies can the UN system deploy to close the connectivity gap, particularly in rural areas?
Despite decades of WSIS progress, a significant portion of the global population remains unconnected. This gap demands further research into scalable, cost-effective models for last-mile connectivity, including the role of postal infrastructure.
How can post offices be better integrated into national digital strategies to serve as inclusive access points, and what investment is needed to connect the estimated 100,000 post offices that currently lack meaningful internet connectivity?
The UPU DG highlighted a concrete gap of approximately 100,000 unconnected post offices worldwide. Further research is needed on financing models, partnership frameworks, and technical solutions to address this infrastructure deficit.
How can AI-enabled health tools be governed responsibly to ensure they are safe, ethical, evidence-based, and equitable across different national contexts?
WHO's DG raised this as a critical challenge. As AI rapidly transforms health systems, there is an urgent need for further research into governance frameworks, standards, and accountability mechanisms that work across diverse health system contexts.
How can the digital divide and the health divide be addressed simultaneously, given that gaps in connectivity, capacity, governance, and financing risk widening health inequalities?
Dr. Tedros explicitly stated that 'the digital divide is a health divide.' This intersection requires further interdisciplinary research to understand how digital exclusion translates into health outcome disparities and what integrated policy responses are needed.
How can AI models for climate and environmental monitoring be made more reliable and accessible to developing countries, given that AI is only as good as the underlying data?
The WMO Secretary General highlighted the fundamental dependency of AI on quality observational data. Further research is needed on how to expand data collection infrastructure in underserved regions and ensure equitable access to AI-driven climate tools.
How can open data standards and open-source software be promoted more broadly across the UN system to replicate the success of WMO's WIS2 model in other domains?
WMO's experience with its unified data policy and WIS2 system offers a replicable model. Further research is needed on how this approach to open standards and data sharing can be adapted and scaled across other UN thematic areas.
How can digital solutions and AI tools be better deployed to support road safety, digital trade standards, and environmental monitoring, particularly in developing countries?
The UNECE Executive Secretary pointed to several practical application areas but noted that much work remains. Further research is needed on how regulatory frameworks for automated vehicles and digital trade can be harmonised globally and implemented in lower-capacity contexts.
How can skills and capacity-building programmes be scaled to ensure all countries can process and benefit from digital solutions, thereby reducing inequality?
Multiple speakers identified the capacity divide as a critical barrier. Further research is needed on effective, scalable models for building digital and AI governance skills across governments, civil society, and the workforce, particularly in the Global South.
What does 'being online' actually deliver in terms of economic value — jobs, sales, income — for people in developing countries, and how can the UN system better support this transition from access to meaningful use?
UNCTAD's Acting Secretary General framed the next decade of digital development around economic value creation. Further research is needed on the enabling conditions — payment systems, logistics, legal frameworks, skills — that determine whether internet access translates into economic opportunity.
How can the WSIS mandate and the Global Digital Compact be better aligned in practice, and what does the joint implementation roadmap mean for concrete country-level action?
UNCTAD tabled a joint implementation roadmap to align WSIS and the Global Digital Compact, but the practical implications for member states remain underexplored. Further research and stakeholder consultation are needed to translate this alignment into measurable outcomes.
How can capacity-building be elevated from an 'add-on' to a strategic priority within the WSIS Plus 20 framework, and what financing mechanisms are needed to achieve this?
UNITAR's Executive Director explicitly noted that capacity building is often treated as an afterthought. Further research is needed on how to institutionalise and fund capacity development as a core component of digital transformation strategies at the national and international levels.
Beyond the digital divide, how should the 'capacity divide' identified in the Independent International Scientific Panel on AI report be measured and addressed across different country contexts?
The newly identified concept of a 'capacity divide' — distinct from mere access — requires further research to define metrics, assess its scale, and develop targeted interventions that go beyond connectivity to address governance, leadership, and institutional capabilities.
How can AI governance frameworks ensure that social justice and inclusion are embedded from the outset, rather than treated as afterthoughts?
UNRISD's Director raised this as a fundamental research and policy question. Further investigation is needed into how AI governance processes can be designed to systematically include marginalised voices and measure impacts on inequality within and between countries.
Who benefits and who is left behind by AI deployment in economies, public services, and daily life, and what policies and institutions are needed to ensure AI promotes sustainable and inclusive development?
These three questions were explicitly posed by UNRISD as the core of their research agenda. They represent significant areas for further empirical investigation, particularly regarding AI's impact on jobs, trust, and well-being across different population groups.
How can the macroeconomic impact of AI on labour markets be better understood and managed, particularly in terms of occupational exposure and workforce transitions?
The ILO Assistant Director General noted ongoing collaborative research with the World Bank on this topic but indicated it requires scaling up. Further research is needed to understand how AI-driven labour market changes affect different sectors, regions, and demographic groups.
How can international labour standards, including the newly adopted treaty on decent work in the platform economy, be effectively implemented across different national contexts with the support of other UN agencies?
The ILO highlighted the adoption of a landmark international treaty but noted the need for cross-agency collaboration in implementation. Further research is needed on enforcement mechanisms, monitoring frameworks, and the role of digital tools in supporting compliance.
How can UNESCO's AI ethics recommendation, adopted by 193 member states, be effectively implemented at the national level, particularly regarding bias, transparency, and accountability?
UNESCO's Assistant Director General noted that implementation support is being provided but the scale of the challenge — 193 member states with varying capacities — requires further research into effective implementation pathways and monitoring mechanisms.
How can digital governance tools such as the data governance toolkit and Internet Universality Indicators be better integrated into national digital strategies to ensure rights-based, open, accessible, and inclusive digital ecosystems?
UNESCO highlighted these scalable tools but their uptake and effectiveness across diverse national contexts requires further evaluation. Research is needed on how to adapt these frameworks to different governance environments and measure their impact.
How can the UN system move beyond fragmented agency-specific digital tools towards shared, multi-tenant, interoperable infrastructure that reduces duplication and costs?
Both UN Ops and UNICC raised this as a critical operational challenge. Further research and planning are needed to identify which systems are most amenable to shared infrastructure approaches and how governance of such shared systems should be structured.
How can digital transformation and AI deployment be designed to reach post-conflict states, least developed countries, and small island developing states without creating dependency on external providers?
UN Ops highlighted the risk that the digital agenda moves fastest for those already connected, leaving the hardest-to-reach further behind. Further research is needed on context-specific models for digital inclusion in fragile and remote settings.
How can the UN system's expertise-on-demand platform, currently being developed by UNDP and ITU, be designed to effectively connect supply and demand for digital and AI expertise across the system?
UNDP's Chief Digital Officer mentioned this as an initiative under the UN80 reform process. Further research and stakeholder consultation are needed to ensure the platform is designed to meet real needs and avoids replicating existing fragmentation.
How should the impact of digital transformation initiatives be measured not only by scale and reach but also through an equity lens, particularly for the most vulnerable populations such as children?
UNICEF's Global CIO raised this as a critical methodological question. Further research is needed on equity-sensitive metrics and evaluation frameworks that can assess whether digital and AI initiatives are reducing or widening disparities for vulnerable groups.
How can AI benchmarks and evaluation frameworks be redesigned to account for child-centric use cases, given that most current benchmarks are designed for adults despite children comprising one third of internet users?
UNICEF identified a significant gap in AI evaluation methodology. This is an emerging area requiring interdisciplinary research involving child development experts, AI researchers, and child rights advocates to develop appropriate assessment frameworks.
How can AI be used responsibly for climate emergency preparedness specifically affecting children, and how can the predictive models be made more accurate and actionable for humanitarian response?
UNICEF's 'Ahead of the Storm' initiative addresses a critical gap in child-focused climate risk modelling. Further research is needed on improving the accuracy of AI-driven hazard forecasts and translating them into operational decision-making tools for humanitarian actors.
How can human rights safeguards be systematically built into AI systems used by the UN and member states, and how can AI be used to strengthen human rights monitoring in resource-constrained and conflict-affected environments?
OHCHR's Chief of Innovation raised this as both a governance and operational challenge. Further research is needed on technical architectures, legal frameworks, and operational protocols that embed human rights principles into AI systems by design.
How can the WSIS action lines be updated to systematically incorporate human rights considerations across all thematic areas, as requested by the UN General Assembly?
The moderator noted that the UN General Assembly resolution asked the WSIS process to include human rights in all action lines. Further research and consultation are needed to determine how this can be operationalised in the roadmaps of each action line facilitator.
How can the UN Pension Fund's model of responsible AI adoption, including ISO 42001 certification, be replicated across other UN entities to ensure transparent and accountable AI governance system-wide?
The Pension Fund's achievement of the first ISO certificate on responsible AI use represents a potential model for the broader UN system. Further research is needed on how this certification framework can be adapted and scaled across diverse UN operational contexts.
How can the UN system develop a shared AI governance framework — encompassing an AI factory, AI gateway, and AI studio — to move AI solutions safely from experimentation to deployment at scale?
UNICC outlined a conceptual three-pillar framework for shared AI governance but acknowledged this is still being developed. Further research and stakeholder engagement are needed to design, test, and implement this framework across the UN system.
How can AI tools for tackling online child sexual exploitation and abuse be made more widely accessible to law enforcement agencies in developing countries, and how can their effectiveness be systematically evaluated?
UNICRI's Centre for AI and Robotics has developed a hub providing AI tools to law enforcement across 129 countries. Further research is needed on expanding access, ensuring tools are adapted to different legal and operational contexts, and measuring their impact on child protection outcomes.
How can industrial AI be made accessible to SMEs in the Global South, and what governance frameworks are needed to ensure responsible adoption in manufacturing and industry?
UNIDO's Chief of Innovation highlighted that responsible AI must reach factories, firms, and workers, not only laboratories and platforms. Further research is needed on the specific barriers SMEs face in adopting industrial AI and on the design of effective support mechanisms.
How can financing for digital development be structured to address the gaps identified in the WSIS Plus 20 review, and what are the implications of different financing models for developing countries?
The ITU Deputy Secretary General noted that UNGIS has been tasked with additional work on financing for digital development as part of the WSIS Plus 20 mandate. This represents a significant area for further research into public and private financing mechanisms, blended finance, and development bank engagement.
With only 35% of SDG targets on track and 2030 approaching rapidly, how can digital technologies and AI be leveraged to accelerate progress across multiple SDG areas simultaneously, including through leapfrogging strategies?
The ITU Deputy Secretary General highlighted the alarming state of SDG progress and the potential role of technology in accelerating it. Further research is needed on which digital interventions have the greatest multiplier effects across SDGs and how these can be prioritised and scaled.
How can AI-enabled adult learning be made more effective, and what does research with Swiss think tanks and the UN International Computing Center reveal about best practices in this area?
UNITAR mentioned a forthcoming collaboration on AI-enabled adult learning as an area of research. This represents an emerging field where further investigation is needed to understand how AI can personalise and scale learning for adult populations, particularly in the Global South.
