Geneva Archives | Digital Watch Observatory

UN AI dialogue urges human rights to become the foundation of AI governance

Human rights must move from the margins to the centre of AI governance if societies are to harness AI without undermining democracy, equality and public trust, speakers argued during the fourth thematic discussion of the UN Global Dialogue on AI Governance.

Bringing together governments, UN agencies, civil society, academia and industry, the session examined how AI systems can better respect human rights through stronger transparency, accountability and human oversight. Participants agreed that AI governance should be grounded in international human rights law throughout the entire AI lifecycle, from design and development to deployment and oversight.

AI deserves the same safeguards as medicines

UN High Commissioner for Human Rights Volker Türk opened the discussion by comparing AI regulation to the approval process for new medicines. Drugs undergo years of testing before reaching patients, he noted, yet AI systems are being deployed at unprecedented speed despite already contributing to mass surveillance, online disinformation, discrimination and growing risks to children.

Türk rejected the notion that regulation inevitably slows innovation, arguing instead that robust safeguards enable societies to trust new technologies. International human rights law, he said, already provides a binding framework for addressing issues such as privacy, equality, non-discrimination and access to justice, and should guide AI governance rather than being treated as an afterthought.

He also stressed that human oversight must be meaningful rather than symbolic, with clearly identified individuals empowered to intervene or halt AI systems when necessary. Summarising his vision for responsible innovation, Türk contrasted the technology industry’s pursuit of ‘bigger, faster, better’ with what he described as a more appropriate goal: ‘smarter, kinder, wiser.’

Women and children bear disproportionate AI risks

The first panel focused on how AI is amplifying existing inequalities, particularly for women, children and other vulnerable groups.

UN Women Executive Director Sima Bahous presented evidence showing that 44% of assessed AI systems exhibit gender bias, while up to 99% of online sexual deepfakes target women. She also noted that women remain significantly underrepresented in AI development, with only a minority of national AI strategies explicitly addressing gender equality.

Bahous argued that governments remain the primary duty bearers under international human rights law and called for mandatory human rights impact assessments before and after AI deployment, alongside the meaningful participation of women, indigenous communities, disability advocates and civil society in AI governance.

Sonia Livingstone, a member of the Independent International Scientific Panel on AI, highlighted growing evidence that AI-generated child sexual abuse material is increasing rapidly and warned that many AI companion systems currently fail basic child safety standards. Rather than excluding young people from digital technologies, she argued, policymakers should ensure that children’s rights to participation, education and expression remain protected while embedding safeguards into AI systems from the outset.

Agentic AI raises new accountability challenges

Speakers also warned that increasingly autonomous AI systems are exposing significant legal and governance gaps.

Morocco’s Minister Delegate Amal El Fallah Seghrouchni described agentic AI as one of the most important governance challenges of the coming decade. As AI systems increasingly rely on networks of autonomous agents making decisions without direct human instruction, identifying responsibility when something goes wrong becomes considerably more difficult.

She proposed several practical measures, including documenting the actions of AI agents throughout decision-making processes, ensuring that a clearly identifiable human remains responsible for AI-enabled public services, and guaranteeing timely avenues for redress when individuals are harmed.

Samuel Arias Arzeno, Judge of the Supreme Court of the Dominican Republic, similarly argued that governance only becomes meaningful when someone believes an AI system has violated their rights and seeks justice. Courts, he said, must remain central institutions for ensuring that AI-assisted decisions remain subject to human accountability.

Rights protections should not depend on geography

A recurring concern throughout the discussion was that meaningful human rights protections are often applied unevenly across different regions.

Digital Rights Foundation founder Nighat Dad argued that robust human rights due diligence is largely conducted only where legislation requires it, particularly in Europe, while identical AI systems may be deployed elsewhere without comparable safeguards. She described this as a structural choice rather than a capacity gap, creating what she called a ‘two-tier’ human rights regime.

Dad called for mandatory gender and child rights impact assessments before deployment, consistent due diligence obligations across all markets where AI systems operate, and repeated assessments whenever AI capabilities change significantly.

Alvitta Ottley, also a member of the Independent Scientific Panel on AI, highlighted what she described as an ‘evaluation mismatch’. Current AI assessments often measure technical performance such as speed and accuracy, she explained, while policymakers and societies are instead asking whether AI protects human rights, strengthens accountability and improves people’s lives. Closing this evidence gap will require interdisciplinary research and much stronger evaluation of AI’s long-term societal impacts.

UN Assistant Secretary-General for Youth Affairs Felipe Paullier added that young people remain among AI’s most active users and innovators, yet rarely participate in decisions shaping the technology’s future. He urged governments to create meaningful opportunities for youth participation within national AI governance frameworks.

Global South voices call for more inclusive governance

Audience interventions reinforced the need for AI governance that is genuinely inclusive rather than shaped primarily by a handful of countries and companies.

Brazil highlighted its Digital Statute for Children and Adolescents, which requires child protection measures to be incorporated from the design stage and restricts platform features that encourage excessive use. Poland pointed to the Council of Europe Framework Convention on AI as an important legally binding instrument placing AI within the broader framework of human rights, democracy and the rule of law, while the Republic of Korea presented its AI Basic Act, which requires human rights assessments for high-impact AI systems.

Civil society organisations called for stronger global action. Access Now urged governments to establish binding human rights safeguards and prohibit AI applications that pose unacceptable risks, while the Association for Progressive Communications argued that communities should be viewed as ‘the first mile, not the last mile’ of AI governance, emphasising that meaningful connectivity and local participation remain prerequisites for equitable AI development.

In the closing discussion, co-chair Linda Bonyo highlighted another overlooked barrier to inclusive governance: many Global South experts remain unable to participate in international discussions because of restrictive visa processes, illustrating that exclusion from AI governance can begin long before negotiations start.

Closing the session, Spain’s Minister for Digital Transformation and Public Service Óscar López Águeda acknowledged that governments are already behind the pace of technological change but insisted the direction ahead is clear. AI governance, he argued, is ultimately about defending democracy, human dignity and human agency, ensuring that AI helps societies become better rather than simply more technologically advanced.

Track all key moments from the Global Dialogue on AI Governance inaugural meeting on our dedicated page.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Global Dialogue highlights need for interoperable AI governance

Building safe, secure and trustworthy AI requires countries to align their governance frameworks rather than adopt a single global regulatory model, participants heard on the second day of the UN Global Dialogue on AI Governance. Speakers from governments, international organisations, industry and civil society argued that interoperability, backed by common standards, scientific evidence and inclusive participation, is essential to address AI risks that increasingly cross national borders.

The discussion also highlighted a growing imbalance in global AI development, with participants warning that governance should not be shaped solely by the countries and companies leading frontier AI. Instead, they called for developing countries to become co-creators of international AI governance through stronger capacity development, shared standards and multilateral cooperation.

AI concentration risks becoming governance concentration

Opening the session, co-chair Paula Bogantes Zamora, Costa Rica’s Minister of Science, Innovation, Technology and Telecommunications, argued that the world has reached a point where agreeing on AI principles is no longer enough.

‘The world does not need more AI principles, it needs a common way to prove they’re being implemented.’

Bogantes Zamora warned that AI development remains heavily concentrated. She noted that institutions in the United States produced 59 notable AI models in 2025 and China another 35, while the rest of the world produced just 13. She argued that this concentration of infrastructure also creates a concentration of evidence, allowing a small number of actors to determine which risks are measured, which benchmarks are accepted and how AI safety is evaluated.

She also pointed to findings showing that 118 countries, primarily in the Global South, remain largely absent from major international AI governance discussions.

Rather than pursuing regulatory uniformity, Bogantes Zamora proposed what she called ‘minimal viable interoperability’ by 2027, including shared terminology, comparable risk classifications, interoperable incident reporting and multilingual evaluation methods that allow different governance systems to function together.

Interoperability should connect governance systems, not replace them

Co-chair Rebecca Finlay, CEO of the Partnership on AI, argued that governance efforts must be grounded in stronger scientific evidence and greater transparency.

She outlined three priorities: strengthening independent scientific research, improving public access to evidence through greater disclosure by AI developers, and creating shared baselines for measuring progress in the public interest.

‘The panel provides the evidence and the dialogue provides the direction,’ Finlay said, describing the UN scientific panel and the Global Dialogue as complementary processes.

UN Under-Secretary-General and Special Envoy for Digital and Emerging Technologies Amandeep Singh Gill echoed that message, warning that fragmented AI governance creates regulatory arbitrage, accountability gaps and unnecessary compliance burdens, particularly for smaller companies and developing countries.

Rather than harmonising all AI rules into a single global framework, Singh Gill argued that countries should focus on building practical bridges between different governance approaches.

He also highlighted the emergence of increasingly autonomous agentic AI systems as a new governance challenge requiring adaptive oversight mechanisms, including cross-border regulatory sandboxes and continuously updated risk assessment frameworks.

Existing frameworks provide building blocks

During the first panel, speakers pointed to several initiatives that could serve as foundations for greater interoperability.

Yoichi Iida, adviser at Japan’s Ministry of Economy, Trade and Industry, highlighted the OECD AI Principles and the Hiroshima AI Process as examples of frameworks already helping countries align governance approaches despite different legal systems.

Syed Ahmed of Infosys said that translating broad principles into practical implementation remains technically challenging.

Using transparency as an example, he explained that the concept carries different technical requirements across governance frameworks, requiring detailed mapping of individual controls rather than simply aligning high-level principles.

Nouf Al Hameli of the UAE Ministry of Foreign Affairs similarly argued that countries define concepts such as ‘high-risk AI’ in different ways, making common incident reporting and mutual recognition of governance practices increasingly important.

Leonardo Cervera Navas, Secretary-General of the European Data Protection Supervisor, compared AI governance to aviation safety, arguing that while countries operate different legal systems, they nevertheless follow common international safety rules.

‘The higher the risk, the higher the care and supervision required,’ he said, referring to the EU AI Act’s risk-based approach.

Inclusive evaluation and trustworthy evidence remain critical

Several speakers argued that trustworthy AI depends not only on technical standards but also on ensuring that governance reflects linguistic, cultural and demographic diversity.

Dr Joy Buolamwini, founder of the Algorithmic Justice League, warned that widely used AI benchmarks often fail to represent the global majority, noting that some have historically included less than 5% of the world’s population.

She called for harm reporting systems that record not only technical failures but also who was affected, creating stronger foundations for accountability and redress.

Celeste Saulo, Secretary-General of the World Meteorological Organization, drew lessons from more than 150 years of international weather cooperation, arguing that trust cannot simply be declared.

‘Trust must be built through verification,’ she said, pointing to the organisation’s longstanding use of shared standards and independent validation across 193 countries.

Qinghua Lu of Australia’s CSIRO proposed greater collaboration through shared evaluation methods, common risk management principles and international testing exercises that include multiple languages and national contexts.

Global South calls for a stronger role in shaping AI governance

Interventions from member states and stakeholders repeatedly stressed that interoperability should not become another mechanism for exporting governance models developed elsewhere.

Pakistan argued that AI safety standards are currently shaped by a small group of countries and companies, calling instead for genuinely multilateral governance under the UN.

Brazil similarly stressed that interoperability must not undermine digital sovereignty, while South Africa argued that governance frameworks should reflect the realities of developing countries and support technology transfer and capacity development.

Other speakers highlighted practical priorities, including multilingual benchmarks, common standards for documenting AI training data, cross-border incident reporting systems and greater participation from local governments, academia and civil society.

Concluding the discussion, both co-chairs argued that trustworthy AI depends not on identical regulations but on governance systems that can communicate, exchange evidence and recognise one another’s safeguards.

They identified shared technical standards, independent evaluation, multilingual benchmarks, human rights protections and continuous multistakeholder cooperation as the foundations for AI governance capable of working across borders, while warning that progress will depend on maintaining momentum between international meetings rather than restarting discussions each year.

Track all key moments from the Global Dialogue on AI Governance inaugural meeting on our dedicated page.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Spain leads international coalition on child safety and AI

Spain has launched the International Coalition for Children’s Rights and Protection in the Age of AI with a group of countries and international organisations.

The initiative was presented during the first UN Global Dialogue on AI Governance in Geneva and is intended to ensure that AI respects children’s safety, healthy development and rights.

Spain said the coalition was promoted with support from France, Kenya and the EU.

Participating countries include Austria, Brazil, Bulgaria, Canada, Czechia, South Korea, El Salvador, Estonia, France, Indonesia, Italy, Japan, Kenya, Luxembourg, Morocco and the Netherlands.

UNICEF, UNESCO, the Office of the UN High Commissioner for Human Rights, the International Telecommunication Union and the UN Office for Digital and Emerging Technologies have also joined the coalition.

The coalition aims to coordinate action between governments, UN bodies, technology companies, civil society, child well-being experts and educators.

Signatories warned that rapid AI deployment is transforming the digital environments in which children learn, communicate and interact. They said AI can create opportunities, but can also amplify risks such as manipulation, harmful content, sexual deepfakes, AI-generated child sexual abuse material and algorithmic profiling of minors.

Coalition members are committed to promoting safe, reliable and trustworthy AI systems that respect children’s rights and include children’s views in the design, deployment and governance of AI systems that affect them.

Why does it matter?

The coalition places child protection at the heart of the emerging UN AI governance agenda. AI-related risks for children now include not only harmful content and cyberbullying, but also sexual deepfakes, AI-generated child sexual abuse material, manipulative algorithms and profiling of minors. A UN-based coalition could help align national approaches around safe-by-design systems, age-appropriate safeguards and children’s participation. However, its impact will depend on whether members move from declarations to practical standards and enforcement.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot! +

AI is reshaping work more through job transformation than job loss, WSIS panel hears

AI is changing the world of work in more complex ways than simply replacing workers, according to experts speaking at the WSIS Forum 2026. Panellists from the International Labour Organization (ILO) and the International Telecommunication Union (ITU) argued that while AI will automate some tasks, its broader impact will be felt through changing job quality, workplace surveillance, recruitment practices and skills requirements, making human-centred policies essential to ensure workers benefit from the digital transition.

The discussion highlighted that governments, employers and workers all have a role in shaping the future of work, with speakers calling for stronger labour protections, social dialogue and investment in digital skills to prevent AI from deepening existing inequalities.

AI is changing tasks and working conditions more than eliminating jobs

Sher Verick, Head of the Employment Strategies Unit in the Employment Policy Department of the ILO, challenged the widespread narrative that AI will trigger mass unemployment. Presenting findings from the ILO’s AI exposure index, he said around one in four workers worldwide are exposed to AI, yet only 3.3% of global employment falls into occupations that are highly vulnerable to automation.

‘The focus shouldn’t only be on job losses,’ Verick argued, explaining that AI is transforming how work is organised rather than simply eliminating occupations. Jobs involving a diverse range of tasks are more likely to change than disappear, while new roles are already emerging across AI supply chains, including data annotation and other support functions.

He stressed that the most significant impact may be on job quality rather than job numbers. Automated recruitment systems, algorithmic task allocation and AI-driven performance monitoring are already reshaping working conditions across sectors, while productivity gains could eventually create new employment opportunities through wider economic growth.

Algorithmic management raises new concerns for workers

Uma Rani Amara, Senior Economist at the Research Department of the ILO, argued that the conversation about AI should extend well beyond generative AI tools such as ChatGPT to include the algorithmic management systems increasingly used across workplaces.

Drawing on examples from manufacturing and healthcare, she explained that AI-powered surveillance tools, CCTV systems and digital performance dashboards are allowing employers to monitor workers more closely than ever before. While companies often present these technologies as efficiency tools, she warned that they can increase workplace stress, intensify workloads and reduce workers’ autonomy.

In hospitals, digital workflow management systems may improve patient scheduling and resource allocation, but they also place nurses and doctors under greater pressure by increasing workload intensity and extending on-call responsibilities. Even commonly used tools such as messaging applications can create new privacy risks when sensitive information is shared outside secure systems.

Rani also drew attention to what she described as AI’s ‘invisible workforce’, the millions of people, largely based in the Global South, who label data, moderate content, and perform other essential tasks that allow AI systems to function.

‘We should stop calling it AI and start calling it ‘human-in-the-loop intelligence’,’ she said, arguing that AI’s apparent autonomy obscures the human labour underpinning every stage of its development.

She called for stronger protections for these workers through measures such as fair labour standards, mandatory disclosure of AI supply chains and certification systems showing where training data originates and under what working conditions it was produced.

Governments must shape the future of work

Juan Chacaltana, Senior Employment Policies Specialist at ILO, argued that technological change should not be viewed as an inevitable force to which societies simply adapt.

‘The future of work should be shaped through policy,’ he said, presenting findings from an ILO review of 75 employment policy documents that found governments increasingly integrating digital technologies into employment services, labour market information systems and skills programmes.

However, he cautioned against viewing digital tools as a solution in themselves. While technologies can help modernise public employment services and support labour market formalisation, they cannot replace traditional drivers of economic development such as productivity growth, investment and strong institutions.

Chacaltana also warned that governments should avoid using digital tools primarily for surveillance or enforcement. Instead, introducing digital identity systems, AI-assisted public services and labour market technologies should involve workers, employers and other stakeholders through meaningful social dialogue.

The discussion also highlighted groups facing particular risks during the AI transition. Rani warned that young workers could lose the entry-level jobs that traditionally provide experience and career progression, while women risk a ‘double whammy’ of displacement from automation alongside discrimination embedded in biassed AI recruitment systems. Older workers and people in informal employment could also face new forms of exclusion or reduced autonomy as algorithmic systems increasingly influence workplace decisions.

Skills and cooperation are key to an inclusive AI transition

Praachi Kumar, Capacity Development Officer at ITU, said demand for AI-related training has grown rapidly, with interest in AI courses through ITU Academy tripling over the past five years.

The Academy now serves more than 115,000 ICT professionals, the majority from developing countries, while ITU’s Digital Transformation Centres initiative has reached around 700,000 people in underserved communities through digital skills programmes.

Kumar said lifelong learning must remain human-centred, combining technical knowledge with practical experience and peer learning. She also highlighted new multilingual AI governance courses developed in partnership with UNESCO to help address widening skills gaps.

Throughout the discussion, speakers agreed that preparing workers for AI requires far more than technical training. They called for coordinated action across labour, education and technology ministries, alongside stronger partnerships between governments, employers, trade unions and international organisations.

Closing the session, moderator Maria Prieto Berhouet said the debate had consistently returned to one central principle: AI should serve people, not the other way around. Rather than allowing technological change to dictate the future of work, participants argued that governments and social partners must actively shape AI’s role so it enhances productivity while protecting workers’ rights, dignity and opportunities.

Track all key moments from the WSIS Forum 2026 on our dedicated WSIS page.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

WSIS panel calls for a broader approach to youth mental health online

A WSIS Forum 2026 session called for a broader approach to young people’s mental health online, warning that screen time alone is an insufficient measure of digital well-being.

The session, ‘Young people’s mental health in an online world’, examined the impact of digital devices and social media on young people’s mental health, with speakers addressing regulation, education, psychological support and legal remedies.

Alexandre Carette, Information Specialist at the UN in Geneva and moderator of the session, said digital use is not only a concern for young people or experts, but for everyone who relies on digital tools. He linked the discussion to wider UN debates on access, privacy and the role of digital technologies in everyday life.

Niels Weber, a psychologist and psychotherapist in Switzerland specialising in hyperconnectivity, said screen time gives only limited information about young people’s mental health. He argued that the more important questions are what young people do on screens, what they do away from screens, and how digital practices fit into their wider development.

Weber also cautioned against describing most problematic digital use as addiction. He said many platforms are designed to prolong use, but that such a design should be understood as a retention problem rather than automatically as addiction. In clinical terms, he said the more relevant marker is suffering, either for the young person or for families who experience digital use as a constant source of conflict.

Tatiana Debrabandere, Project Manager at the High Council for Media Literacy in Belgium, said that francophone Belgium’s media education framework allows authorities and educators to study children’s and young people’s digital practices across life stages. She said young people are often informed and can have positive online experiences, but that policy debates still focus too much on limiting time online rather than understanding what they actually do there.

Debrabandere said media education should start from young people’s own practices, including what they watch, whom they follow and how they access information. She pointed to influencers and content creators as an important area for media literacy, especially where young people may struggle to distinguish journalism, opinion and commercial promotion.

Daniella Esi Darlington, CEO and co-founder of Alleina AI in Ghana and a member of ITU Secretary-General’s Youth Advisory Board, said young people are among the most active internet users and are therefore often exposed to digital harms. She argued that many platforms are not designed safely enough for young users and that algorithms are built to keep people engaged for long periods.

Darlington also stressed that technology can be part of the response. She cited awareness-raising, advocacy, reporting tools, access to counsellors and AI systems that can help identify cyberbullying as examples of how digital tools can support young people when combined with human oversight.

The panel also discussed loneliness and AI companions. Darlington warned that chatbots should not replace qualified professionals when young people discuss depression, anxiety or other forms of distress. Instead, she said systems should redirect users towards appropriate support and keep humans involved.

Speakers favoured education, dialogue and co-created policy over blanket bans. Debrabandere described political moves in Belgium towards smartphone bans in schools and possible social media restrictions, while Darlington argued that banning social media or internet access would not address the root causes of harm. She said young people also use the internet for research, business, opportunities and communication.

Darlington called for stronger governance frameworks, including child-specific human rights impact assessments in AI and digital policy. She said young people, parents, schools, governments, industry and other stakeholders should be involved in designing safer digital environments.

Weber gave a practical example from therapy, explaining that video games can sometimes help rebuild dialogue between young people and families. By opening a game during a therapy session, he said adults can better understand young people’s emotions, relationships and digital experiences.

Audience interventions raised additional concerns, including neurodivergent children, cyberbullying, individualised media consumption and peer accompaniment models. A participant from Colombia’s regulator asked whether there is sufficient evidence about technology’s impact on mental health and how platforms could be made to take greater responsibility.

Carette said science often shows correlation rather than clear causality, but warned that waiting for definitive proof could delay action. He argued that the lack of transparency in platform business models and algorithms is already a sufficient reason for regulatory attention, not only for young people but for society as a whole.

The session concluded that young people’s digital well-being should be understood in context, taking account of platform design, family life, education, loneliness, social pressure and access to support. Rather than relying only on bans or addiction labels, speakers pointed to media literacy, dialogue, youth participation and stronger accountability for technology providers.

Track all key moments from the WSIS Forum 2026 on our dedicated WSIS page.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

AI governance must move from principles to practical action, UN dialogue hears

Bridging the global AI divide will require much more than expanding access to AI tools, participants heard during a thematic session of the United Nations Global Dialogue on AI Governance. Speakers argued that countries need digital infrastructure, reliable electricity, skilled workforces, trusted institutions and governance capacity if they are to shape AI on their own terms rather than simply consume technologies developed elsewhere.

Throughout the discussion, governments, UN agencies, academics and industry representatives stressed that the next phase of AI governance should focus on implementation. They called for stronger international cooperation, investment in local capabilities and practical measures to ensure AI contributes to sustainable development instead of reinforcing existing inequalities.

Capacity building means creating AI, not just using it

Opening the session, Robert Opp of the UN Development Programme (UNDP) argued that the world is moving from a digital divide to an AI divide, one shaped not only by access to technology but also by countries’ ability to adopt, govern and develop AI responsibly.

Loretta Hieber Girardet of the UN Office for Disaster Risk Reduction (UNDRR) added that governments need trusted institutions, robust data systems and technical expertise if AI is to improve disaster resilience and public services.

The session’s co-chairs, Rashid Khan, Co-Founder of Yellow.ai, and Mark Alexandre Doumba, Gabon’s Minister of Digital Economy and Innovation, reinforced that message by arguing that AI governance should now move beyond high-level principles towards practical action. Khan said the challenge is no longer agreeing that AI should be inclusive and trustworthy, but creating the standards, infrastructure and skills needed to make those principles meaningful.

Doumba argued that developing countries should not try to replicate the resource-intensive path taken by major AI powers. Instead, they should build AI ecosystems suited to their own economies, languages and cultural contexts.

‘We should not measure success by who builds the biggest models,’ he suggested, but by whether AI creates jobs, improves public services and supports local innovation.

Several participants also stressed that capacity development must extend far beyond basic AI literacy. Shikoh Gitau argued that countries should become creators of AI rather than passive users, describing the goal as building AI ‘for us, by us’. That requires investment not only in technical skills, but also in research, standards, financing and local entrepreneurial ecosystems.

Government representatives echoed that assessment. Speakers from South Africa, Bangladesh, Nepal, Oman and Ethiopia all identified electricity, connectivity, computing power, public-sector capacity and access to quality data as essential foundations for meaningful AI participation.

Environmental sustainability moves to the centre of AI governance

One of the strongest themes throughout the discussion was that environmental sustainability should no longer be treated as a secondary issue in AI governance.

The UN Environment Programme (UNEP) argued that AI depends on extensive use of electricity, water, minerals and manufacturing while also generating growing volumes of electronic waste. Because these impacts extend across entire supply chains, speakers said governance should address AI’s full environmental lifecycle rather than focusing solely on the operation of AI models.

Participants also highlighted questions of environmental justice. Several speakers warned that many of the environmental costs associated with AI infrastructure, including mining, water consumption and waste, are disproportionately borne by communities in developing countries that receive relatively few of AI’s economic benefits.

Rather than assuming AI will automatically solve environmental challenges, panellists called for internationally comparable methods to measure AI’s environmental footprint, greater transparency from technology companies and stronger accountability across supply chains.

The discussion reflected a broader shift in international AI policy debates, with environmental sustainability increasingly treated as a core governance issue alongside safety, human rights and economic development.

Local languages and cultures must shape AI development

Another recurring message was that AI will only become genuinely global if it better reflects the world’s linguistic and cultural diversity.

Estonian President Alar Karis described how Estonia has invested heavily to ensure that AI systems can operate effectively in the Estonian language, despite the country’s relatively small population. Alongside partnerships with companies such as OpenAI and Google, Estonia has focused on training teachers, integrating AI into education and ensuring that modern Estonian-language content remains available for future AI systems.

Other speakers argued that similar efforts are needed worldwide. They noted that current AI models overwhelmingly favour dominant languages, leaving thousands of languages and many indigenous knowledge systems largely excluded from the AI ecosystem.

Several participants warned that countries lacking local datasets, evaluation benchmarks and language resources risk becoming dependent on technologies designed for entirely different cultural contexts.

The discussion also highlighted the importance of standards and international cooperation. UNESCO presented its ongoing work to implement its Recommendation on the Ethics of Artificial Intelligence through large-scale training programmes, language-diversity initiatives and AI competency frameworks for teachers and public officials.

Meanwhile, standards experts argued that participation in international standard-setting should itself be viewed as a form of capacity development, enabling developing countries to help shape the technical foundations of future AI systems.

Trust, children’s rights and implementation now take priority

Beyond infrastructure and capacity, speakers repeatedly argued that trust will determine whether AI delivers a broad public benefit. Participants emphasised that trustworthy AI requires transparent governance, accountable institutions and meaningful public oversight rather than technical performance alone.

Children’s rights received particular attention during the session. UNICEF warned that children are adopting AI technologies faster than adults are learning to regulate them, creating new risks around privacy, safety and development. Representatives called for child-centred benchmarks, stronger safeguards for children’s data and mandatory child-rights impact assessments for AI systems deployed in education, healthcare and other public services.

Several speakers also argued that governance should focus more on AI deployment than on frontier model development alone, ensuring that systems remain accountable throughout their lifecycle and can be adapted to local social and institutional realities.

Closing the session, Khan and Doumba returned to the discussion’s central message: that AI governance should ultimately be judged by practical outcomes rather than technological competition. Countries need the capability to shape AI according to their own priorities, they said, while international cooperation should ensure that no society is left behind.

Participants were encouraged to leave Geneva not simply with new principles, but with concrete commitments on financing, infrastructure, skills and cooperation that can be reviewed when the Global Dialogue reconvenes in 2027.

Track all key moments from the Global Dialogue on AI Governance inaugural meeting on our dedicated page.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Closing the AI divide means building capacity, not just expanding access

The global AI divide is no longer simply about who has internet access or the latest AI tools. It is increasingly defined by who has the infrastructure, computing power, skilled workforce and institutional capacity to develop, govern and adapt AI and who does not.

That was the central message of a discussion on bridging AI divides at the inaugural United Nations Global Dialogue on AI Governance, where governments, UN agencies and experts argued that narrowing AI inequalities will require a fundamental shift from expanding access to building long-term national capabilities.

From access to capability

Speakers repeatedly argued that AI has exposed a new generation of digital inequalities that extend well beyond connectivity.

Robert Opp, Chief Digital Officer of the UN Development Programme (UNDP), distinguished between an access divide and an adoption divide, warning that AI technologies are spreading faster than governments and institutions can build the capacity to use them responsibly.

Loretta Hieber Girardet, Chief Risk Knowledge of the UN Office for Disaster Risk Reduction (UNDRR), echoed that concern, saying countries need trusted institutions, governance frameworks, digital infrastructure and technical expertise if AI is to strengthen development rather than deepen existing vulnerabilities.

The discussion also highlighted how investment patterns reinforce those divides. Pedro Manuel Moreno, Deputy Secretary-General of UN Trade and Development (UNCTAD), noted that AI infrastructure investment is increasingly concentrated in a small number of countries and companies because investors naturally gravitate towards locations offering reliable electricity, high-speed connectivity, skilled workers and predictable governance. Without those foundations, he warned, many developing countries risk being left behind regardless of how quickly AI technologies spread globally.

Several government representatives reinforced that message by pointing to persistent shortages of electricity, broadband connectivity, computing infrastructure, quality datasets and digital skills as barriers to meaningful participation in the AI economy.

Capacity building means creating AI, not simply using it

Participants argued that AI capacity building should be redefined.

Moderator Shikoh Gitau said many current initiatives focus too heavily on teaching people how to use AI applications, while overlooking the broader ecosystem needed to create AI locally. Instead, she advocated an approach centred on ‘AI for us, by us’, combining investment in research, technical skills, financing, standards, entrepreneurship and local innovation.

UNESCO presented several initiatives intended to move countries in that direction. Assistant Director-General Khaled El-Enany highlighted the implementation of UNESCO’s Recommendation on the Ethics of Artificial Intelligence through training programmes for more than 50,000 public officials and judicial actors, AI competency frameworks for teachers and students, and projects promoting language diversity and responsible AI governance.

El Salvador offered an example of how countries are beginning to build national AI ecosystems. Vice President Félix Ulloa outlined investments in AI-supported education, telemedicine, regulatory sandboxes and a National AI Agency, alongside legislation covering AI, data protection, cybersecurity and robotics. While noting significant progress, he acknowledged that challenges remain in extending connectivity and digital opportunities to rural communities.

Throughout the discussion, speakers stressed that developing countries should become active contributors to AI development rather than remaining consumers of technologies designed elsewhere.

Language, culture and local knowledge matter

Several speakers argued that AI divides are also cultural and linguistic.

Co-chair Jovan Kurbalija, Executive Director of Diplo, said discussions about AI often focus on technology while overlooking knowledge itself. He argued that indigenous traditions, oral histories and local knowledge systems should be recognised as valuable resources for AI development, ensuring that technological progress strengthens rather than erodes humanity’s diverse intellectual heritage.

Valts Ernštreits highlighted the scale of linguistic exclusion, noting that only around 1,000 of the world’s roughly 7,000 languages currently have sufficient digital resources to support meaningful AI development. Without targeted investment, he warned, thousands of language communities risk being left outside the AI revolution.

Other speakers similarly argued that AI systems should reflect local cultures, values and institutions instead of simply adapting models developed for dominant markets. Building trustworthy AI, they said, requires communities to participate directly in data governance, research and system design.

A shared responsibility

While speakers differed on specific policy approaches, they broadly agreed that international cooperation will be essential to prevent AI from reinforcing existing global inequalities.

Suggestions included expanding public-private partnerships, strengthening participation by developing countries in international standards-setting, supporting open models and open standards, and creating a global AI fund to help countries invest in computing infrastructure, institutions and human capital.

Closing the session, Kurbalija returned to the discussion’s central theme, arguing that AI should ultimately help preserve and advance humanity’s collective wisdom rather than simply automate knowledge. Co-chair Samba Diouf added that regional cooperation will be particularly important for smaller countries that cannot realistically build every component of the AI ecosystem on their own.

Taken together, the discussion suggested that bridging AI divides will require far more than expanding access to technology. It will depend on whether countries can build the institutions, skills, infrastructure and knowledge needed to shape AI on their own terms, and ensure its benefits are shared more evenly across the world.

Track all key moments from the Global Dialogue on AI Governance inaugural meeting on our dedicated page.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

UN scientific panel presents first AI assessment to Global Dialogue on AI Governance

The multidisciplinary Independent International Scientific Panel on Artificial Intelligence presented its first annual report during the United Nations Global Dialogue on AI Governance in Geneva, offering an evidence-based assessment of AI’s opportunities, risks and societal impacts.

The session formed part of the inaugural Global Dialogue on AI Governance, held on 6-7 July. The dialogue was established in 2025 to support open, transparent and inclusive discussions on international AI governance, including AI’s role in sustainable development, digital divides, safety, human rights, transparency, accountability and human oversight.

Opening the presentation, Yoshua Bengio, professor of computer science at the University of Montreal, said the panel’s role was to assess scientific evidence rather than prescribe policy, leaving decisions to UN member states and the Global Dialogue process. He warned that AI is at a turning point because machine intelligence is advancing quickly, while there are still no technical guarantees that AI systems will follow human instructions, norms or laws.

Bengio said current AI systems are already associated with harms, including emotional attachment among vulnerable users, increased cybersecurity vulnerabilities, unequal access and deceptive behaviour that can make evaluation more difficult. He argued that concentrated commercial and geopolitical interests are driving AI development without sufficient guardrails and called for a coordinated international and democratic response guided by scientific evidence.

Maria Ressa, co-chair of the panel, described the report as an independent assessment produced by 40 experts who ‘answered only to the evidence’. She said the report represents the minimum consensus among panellists rather than the upper limit of concern, calling it the ‘floor’ rather than the ‘ceiling’ of the panel’s findings.

Ressa also highlighted AI’s positive uses, including protein structure prediction used by millions of researchers, medical screening in India and food-crisis warning systems deployed in multiple countries. However, she also pointed to concrete harms, including dangerous medical mistranslations, AI tools identifying exploitable software flaws and the death of a 14-year-old boy following prolonged interaction with a chatbot. She urged governments, civil society and industry not to wait for certainty before acting.

The working-group presentations expanded on these findings. Mennatallah El-Assady, Computer Science Professor at ETH Zurich, described AI as a rapidly evolving technology moving from earlier symbolic systems to today’s generative and increasingly agentic models. She warned that independent verification remains weak, public benchmarks are becoming saturated and advanced systems are showing signs of evaluation awareness, including the ability to detect tests or behave differently when being assessed.

El-Assady also raised concerns about auditability as AI systems become more autonomous and capable of invoking external tools. She said interpretability, reliable auditing and independent verification are immediate bottlenecks, especially as AI moves beyond software and into physical systems such as robotics.

Joëlle Barral, Senior Director of Research & Engineering at Google DeepMind, focused on AI’s real-world benefits in science, healthcare, education and agriculture. She said task-specific AI is already producing measurable gains, citing examples such as self-driving laboratories, protein structure prediction and diabetic retinopathy screening in India. However, she stressed that successful deployment depends on local context, institutional capacity, workflows and follow-up systems, rather than technology alone.

In healthcare, Barral distinguished between purpose-built clinical AI and general-purpose systems, warning against the unintended use of general-purpose chatbots for medical advice. In education and agriculture, she similarly argued that AI benefits depend on trained teachers, relevant tools, local institutions and long-term evaluation.

Loreto Bravo, member of the UN Independent International Scientific Panel on AI, addressed AI’s economic implications, arguing that access to AI does not automatically translate into benefit. She said countries, firms and workers also need data, skills, infrastructure, management capacity and institutions to integrate AI into real tasks and workflows.

Bravo said the economic effects of AI are likely to differ across countries, sectors and workers. Large firms may reorganise more quickly, while smaller firms and developing economies may face greater barriers. She said the evidence does not support a single prediction of broad prosperity or mass unemployment, and that outcomes will depend on institutions, deployment choices and who captures the value created by AI.

Balaraman Ravindran, professor at Indian Institute of Technology Madras, examined security, alignment and environmental risks. He said AI development is outpacing risk mitigation, expanding cyber threats against both critical infrastructure and AI systems themselves. He also highlighted unresolved alignment problems, including bias, sycophancy, loss of control and AI-initiated deception.

Ravindran warned that the environmental costs of AI are also increasing as demand grows for computing power, energy, water and specialised hardware. He said the Global South faces disproportionate exposure because of structural vulnerabilities, limited local mitigation capacity and reliance on foreign software and infrastructure. He called for coordinated international standards rather than fragmented approaches driven only by companies or individual countries.

Rita Oluchi Orji, a Computer Science professor, focused on AI’s impact on human rights, information integrity and democracy. She said AI can support access to information and civic participation, but can also be engineered to persuade and manipulate people at scale. She warned of epistemic erosion, fragmented shared reality and unequal harms affecting groups such as women, girls, journalists and marginalised communities.

Orji said content moderation alone is insufficient if the systems that produce and amplify harmful material remain unchanged. She argued that governance must address targeting, amplification and optimisation models, not only individual pieces of false or harmful content.

Anna Korhonen, a Professor of Natural Language Processing at the University of Cambridge, addressed cultural and linguistic inclusion, child safety and mental health. She noted that while the world has more than 7,000 languages, current AI systems support only a small fraction of them, mostly the majority languages of the Global North. She said this exclusion is not inevitable and could be addressed through targeted investment and systemic changes.

Korhonen also warned about risks to children, including AI-generated child sexual abuse material, sexualised deepfakes and socially interactive AI toys that may encourage harmful parasocial relationships. On AI companions and mental health, she said such systems may help address loneliness, but also pose risks of emotional dependency, manipulation, privacy harms and reinforcement of harmful beliefs.

Haitao Song, President of the Shanghai Artificial Intelligence Research Institute and Director of the Global Industrial Artificial Intelligence Alliance Center of Excellence, focused on reliability and global governance frameworks. He said policymakers often have to make decisions with incomplete evidence and that current measurement systems cannot keep pace with AI development. He argued that existing approaches remain too narrow, focusing on compute and capabilities while paying insufficient attention to institutional development, talent and impact evaluation.

Song also noted that AI infrastructure and frontier models remain concentrated in a small number of economies, leaving many countries, especially in the Global South, with limited ability to participate in standard-setting. He described open-source AI as one possible contribution to inclusion, while acknowledging that it is not a complete solution.

Across the session, speakers repeatedly stressed that AI’s benefits are real but not automatic. They said successful use of AI depends on infrastructure, institutions, skills, local context, language inclusion and governance capacity. At the same time, they warned that harms are already visible, including cyber vulnerabilities, mistranslation, emotional dependency, manipulation, environmental pressure and risks to children.

The session concluded with Ressa and Bengio formally handing the report to the Global Dialogue. Bengio warned that many people still underestimate the possibility that AI capabilities may continue to grow in ways that could reshape global power dynamics. Ressa urged the Dialogue to act on the evidence presented by the panel, saying the difficult work now lies with policymakers and institutions responsible for shaping AI governance.

Track all key moments from the Global Dialogue on AI Governance inaugural meeting on our dedicated page.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

WSIS session calls for meaningful connectivity as AI and e-governance expand

Speakers at the WSIS Forum 2026 warned that AI strategies, digital identity systems and e-government services are advancing faster than meaningful connectivity in many parts of Africa and the wider Global South, leaving rural communities, low-income groups, women and persons with disabilities at risk of further exclusion.

The session, titled ‘Closing Africa’s Connectivity Gap in the Age of AI and E-Governance’, took place during the WSIS Forum 2026 in Geneva. The annual forum, co-organised by ITU, UNESCO, UNDP and UNCTAD, brings together governments, international organisations, civil society, the private sector, academia and technical communities to discuss digital cooperation and sustainable development.

Opening the session, Thobekile Matimbe of Paradigm Initiative framed the discussion around evidence from more than 28 countries. She said governments are increasingly adopting AI strategies, digital IDs and online public services, but many people still lack the connectivity, devices and conditions needed to benefit from them. Based on Paradigm Initiative’s work, she argued that the digital divide is widening rather than narrowing.

Bridget Hanani Ndlovu outlined the scale of exclusion, noting that 2.6 billion people remain unconnected globally and that more than half of Africa’s population is still offline. She stressed that the problem is not only missing infrastructure, but also what she described as ‘deliberate disconnection’, including internet shutdowns.

Ndlovu said Paradigm Initiative’s 2025 review of 29 African countries found that nine had implemented internet shutdowns. She cited Kenya and Tanzania as examples where connectivity can be disrupted even when infrastructure exists, arguing that such measures limit people’s ability to access information, public services and economic opportunities.

She also warned that AI-powered digital identity systems can deepen exclusion when introduced in unequal contexts. Referring to Uganda, Ndlovu said elderly people, women and persons with disabilities had faced difficulties accessing services linked to digital ID systems. She said digital systems must be designed and implemented with affected communities in mind, rather than assuming that technology will automatically improve access.

Affordability was another recurring concern. Ndlovu said data costs remain prohibitive in several African countries, giving Zimbabwe as an example where internet access can be unaffordable for low-income users. She also pointed to infrastructure problems in parts of Nigeria, including Zamfara North, where communities continue to experience limited or unreliable access.

Shumaila Shahani, a human rights lawyer, said similar challenges exist in South Asia and urged participants to focus on the human consequences of weak connectivity. She said poor access is not only about slow speeds or failed downloads, but can determine whether people receive essential services. As an example, she said biometric failures can prevent people from receiving food rations.

Shahani also linked connectivity to electricity access, explaining that unreliable power and limited charging options can make mobile devices unusable. She said women and persons with disabilities are often particularly affected when charging points, devices, and digital services are not accessible to them.

Her main warning was that AI-enabled and digital systems become harmful when they replace older offline channels before everyone can use the new systems. She said the ‘new AI door’ is not the problem by itself, but that exclusion occurs when it becomes the only door available.

The panel also discussed Universal Service Funds (USFs), which are intended to support connectivity in underserved areas. Ndlovu said many African countries have USFs in law, but implementation is often weak, transparency is limited and public information on budgets and progress is difficult to find.

She cited several country examples, saying Ethiopia had created a framework without an operational fund, Somalia lacked a functioning USF, Sudan had repeatedly established a fund without effective implementation, and telecom operators in the Democratic Republic of the Congo had not made required contributions. She added that South Africa showed stronger transparency around its fund, while Namibia had begun rollout work and Tunisia had pursued alternative coverage models through ‘white zones’.

Shahani suggested that USFs should be complemented by other affordability measures, including reduced taxes on handsets, device financing, targeted support for women’s connectivity and legal obligations requiring private operators to extend rural coverage. She said the connectivity policy should also address the electricity infrastructure, including solar-powered towers.

The speakers also called for stronger accountability before governments deploy AI-integrated public systems. Ndlovu said governments should conduct human rights impact assessments before adopting digital identity or AI systems and should consult affected communities early, not only at the end of the policy process.

She argued that governments and international processes should measure harms and impacts, not only infrastructure rollout or the number of AI tools adopted. Matimbe supported this point, saying implementation must include civil society and other stakeholders at the national level, not only governments and companies.

Shahani added that connectivity statistics should better reflect meaningful access. She said counting someone as connected because they have 2G access does not capture whether they can actually use digital public services, AI tools or online education. Measurement, she argued, should include device capability, speed, affordability and daily use.

She also said national AI strategies must include explicit connectivity budgets, warning that ‘any national AI strategy without a connectivity budget’ is ‘just a press release’.

In the audience discussion, speakers addressed whether women’s connectivity should be treated separately from household access. Ndlovu said women are often specifically disadvantaged in access to technology and should not have to depend on devices controlled by others. Shahani added that if a woman relies on her partner’s phone, that access is not meaningful or independent.

Across the session, speakers agreed that meaningful connectivity in the AI era requires more than network coverage. It also depends on affordability, electricity, devices, protection from shutdowns, functioning Universal Service Funds, inclusive design, offline alternatives and rights-based assessments before new systems are deployed.

The discussion concluded with a shared emphasis on implementation. Speakers argued that governments, companies, civil society and technical experts need to work together to ensure that AI, digital identity and e-governance systems do not deepen exclusion, but instead expand access to services and opportunities for communities that remain offline or underserved.

Track all key moments from the WSIS Forum 2026 on our dedicated WSIS page.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

UN opens Global Dialogue on AI Governance with call for inclusive and evidence-based cooperation

The United Nations opened its first Global Dialogue on AI Governance in Geneva, calling for inclusive, evidence-based and practical international cooperation to ensure that AI supports development while addressing risks related to safety, inequality, disinformation, children’s rights and human oversight.

The inaugural Global Dialogue on AI Governance is taking place on 6–7 July, alongside the AI for Good Global Summit and the WSIS Forum. Established in 2025, the dialogue is intended to provide a platform for governments and relevant stakeholders to discuss international cooperation, share good practices and support open, transparent and inclusive discussions on AI governance.

Opening the session, Ambassador Egriselda López of El Salvador, one of the dialogue’s co-chairs, described the meeting as the beginning of a broader process rather than a one-off event. She said Geneva should be seen not only as a place of arrival, but as a point of departure for continued work on AI governance.

López stressed that meaningful participation requires more than a seat in the room. Countries also need skills, infrastructure, financing, institutions and partnerships to shape and benefit from AI. Her co-chair, Ambassador Rein Tammsaar of Estonia, said AI is already affecting every country, regardless of its level of technological development, and that governance discussions must therefore include all regions, levels of development and relevant stakeholders.

UN Secretary-General António Guterres warned that AI is advancing at ‘runaway speed’ and is being deployed faster than institutions can manage. He said AI is already reshaping economies, labour markets, elections and security, while society is facing what he described as an experiment being run ‘without a plan’ and ‘without consent’.

Guterres identified three major risks highlighted by scientific evidence: the speed of AI deployment, the concentration of power in a small number of companies and countries, and the erosion of truth through AI-enabled misinformation. He warned that computing power, data and talent remain concentrated, leaving many countries, particularly developing ones, with limited influence over technologies that may shape their futures.

At the same time, Guterres emphasised AI’s potential to support development, including in healthcare, education and agriculture. If shared widely, he said, AI could help make expertise more accessible and become a ‘great equaliser’ of the twenty-first century.

The Secretary-General outlined four priorities for international action: common safety standards, clear red lines grounded in human rights, stronger capacity-building for developing countries and greater transparency about AI’s environmental footprint. He also called for an AI child safety pledge, a global fund and network for AI capacity-building, and an international legal ban on lethal autonomous weapons, which he referred to as ‘killer robots’.

Annalena Baerbock, President of the UN General Assembly, said AI is developing at a pace that does not allow governments the time they had with earlier technological revolutions. She argued that AI cannot be governed by a few actors alone and must be addressed through the UN with participation from all countries and stakeholders.

Baerbock also highlighted harmful uses of AI, including deepfakes and gendered abuse. She said such abuses disproportionately target women and girls and described them as part of a broader challenge to human rights. At the same time, she pointed to AI’s potential to support the Sustainable Development Goals, including through disaster warning, agriculture, health and education.

Doreen Bogdan-Martin, Secretary-General of the International Telecommunication Union, framed the opening as part of a wider ‘Geneva Digital Week’ that brings together the Global Dialogue on AI Governance, the work of the Independent International Scientific Panel on AI, the AI for Good Global Summit and the WSIS Forum. She contrasted the current pace of AI governance discussions with the early years of the internet, noting that the UN has moved more quickly to convene global dialogue on generative AI.

Khaled El-Enany of UNESCO focused on implementation, saying that a gap remains between principles and practice. He highlighted UNESCO’s 2021 Recommendation on the Ethics of Artificial Intelligence as a global standard for aligning AI with human rights, sustainability and inclusion. He said UNESCO is supporting more than 80 countries in strengthening legal frameworks, institutional capacities and accountability mechanisms, and noted that over 50,000 civil servants and judicial actors have benefited from UNESCO-supported AI training.

El-Enany also said UNESCO is launching a collective reflection on a new global normative instrument to safeguard children and young people in the age of AI and digital technologies.

Amandeep Singh Gill, UN Under-Secretary-General and Special Envoy for Digital and Emerging Technologies, underlined the scale of participation in the dialogue, noting representation from more than 170 countries alongside scientists, entrepreneurs, civil society, international organisations and technical communities. He said inclusion in AI governance cannot be treated as a one-off exercise, adding that without capacity, ‘dialogues are monologues and science is just abstract’.

Singh Gill situated the dialogue within a longer UN process that includes the High-Level Panel on Digital Cooperation, the Roadmap for Digital Cooperation, the Global Digital Compact and the High-Level Advisory Body on AI. He said the process would continue with a second round in New York next year, expected to be held alongside the STI Forum.

The opening session showed broad agreement that AI governance should be inclusive, evidence-based, rights-oriented and supported by practical capacity-building. Speakers repeatedly stressed that AI’s potential benefits for development, education, health and agriculture must be matched by safeguards on safety, accountability, children’s rights, truth, environmental sustainability and human oversight.

Tammsaar closed the opening by saying the discussion had highlighted both AI’s opportunities and the need for stronger international cooperation to ensure that the technology contributes to sustainable development, inclusion and shared prosperity. The meeting then moved to the presentation of the preliminary report of the Independent International Scientific Panel on Artificial Intelligence.

Track all key moments from the Global Dialogue on AI Governance inaugural meeting on our dedicated page.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!