High-level governmental plenary segment: Harnessing the benefits of AI for all through inclusive and interoperable approaches -part 3

Germany firmly believes that international cooperation is essential to ensuring AI serves humanity as a whole, given the rapid pace of AI development. The speaker emphasises that without proactive shaping of AI through international collaboration, AI will end up shaping humanity instead.

The speaker stated that AI will develop rapidly, perhaps faster than any of us realise, and that to match this speed, international cooperation is essential . He further warned that 'we must shape AI because otherwise AI will shape us' .

Germany insists that safety and security cannot be afterthoughts added at the end of AI development but must be embedded from the very beginning as a first principle. This includes respect for human values, accessibility for everyone, and ensuring the technology is never directed against humans.

The speaker called for embedding 'safety by design, rules of conduct, and respect for human values' as a first principle, ensuring the technology is accessible to everyone and never directed against humans . He reiterated that 'AI safe and secure by design must be the guiding principle' and that 'safety and security cannot be added at the end' .

Germany is taking concrete national action by establishing a dedicated National AI Safety and Security Institute to evaluate advanced AI systems and advise government. This institute will be connected to both European and international networks, reflecting Germany's commitment to collaborative AI governance.

The speaker announced that Germany will establish a National AI Safety and Security Institute anchored in cutting-edge research and plugged into European and international networks, with the job of evaluating advanced AI systems and advising government in international collaboration .

Germany argues that a multi-stakeholder approach is not optional but a necessity for effective AI governance, requiring meaningful participation from a broad range of actors. This includes not only industry and academia but also those who will feel the direct impact of AI, such as workers, learners, and citizens.

The speaker stated that 'a multi-stakeholder approach is a necessity' and that industry, academia, and civil society, especially from underrepresented regions, must have a meaningful seat at the table . He further specified that this must include workers, learners, entrepreneurs, public administrations, and citizens .

The EU views this global dialogue as a singular opportunity to establish the groundwork for sustained international cooperation on AI governance. The speaker highlights that frontier AI is reaching capabilities that were unimaginable even weeks ago, making this moment particularly critical.

The speaker described the dialogue as 'a unique moment' and stated that 'a successful first global dialogue can lay the foundation for greater international cooperation on AI, which is truly needed in a fast-evolving AI landscape' . He noted that 'today, frontier AI is reaching capabilities that were simply unimaginable before and even a few weeks ago were not imaginable' .

The EU has adopted a governance approach that balances innovation with responsibility, using uniform risk-based rules across its single market that only intervene when risk is too high or unacceptable. This approach is both human-centric and grounded in international human rights standards.

The speaker stated that 'the EU has adopted the approach of responsible and trustworthy AI innovation' and that their approach is 'both human-centric and grounded on international human rights' . He noted that EU legislation 'only intervenes when the risk is too high or unacceptable' .

The EU emphasises that AI governance must be informed by facts and evidence, and that no single actor has all the answers. The EU comes to the dialogue with intellectual honesty, recognising the need for independent scientific panels, multi-stakeholder engagement, and multilateral cooperation.

The speaker stated that 'governance should be informed by facts and evidence' and highlighted the importance of the work of independent international scientific panels on AI . He explicitly said 'we come with intellectual honesty to say we don't know everything' and that 'evidence, multistakeholder dialogue, and multilateralism are needed at this very moment' .

The EU is making substantial investments in AI infrastructure, including upgrading its network of public supercomputers, AI factories, and gigafactories, and expanding Eurocompute capacity. The EU views AI as the defining driver of its future prosperity and competitiveness, promoting an AI-first approach.

The speaker stated that the EU will 'upgrade our unique network of public software computers and AI factories and gigafactories' and is 'expanding Eurocompute capacity, offering this to scientists and startups' . He noted this is supported by the recent AI and Cloud Development Act and that 'AI will be the defining driver of Europe's future prosperity and competitiveness' .

Somalia argues that unlike telecommunications or other digital technologies, AI is inherently global in nature, with its deployment and impact crossing national borders. This makes it impossible for any single country or regional bloc to govern AI effectively, necessitating unprecedented levels of international cooperation.

The speaker stated that 'AI cannot be effectively governed by any single country or even a regional bloc acting alone' and that 'AI is inherently global' with its 'deployment and impact crossing national borders, requiring unprecedented international cooperation' .

Somalia welcomes the global dialogue as an important step towards building a more inclusive global AI governance architecture where all nations can contribute. The speaker emphasises that for countries like Somalia, overcoming fundamental constraints must come before fully harnessing AI's benefits.

The speaker welcomed 'the convenience of this timely and important inaugural global dialogue on AI governance' and commended the UN for 'creating an exclusive platform where all member states can contribute to shaping the future of AI' . He noted that Somalia 'welcomes this global dialogue as an important step towards a more inclusive global AI governance architecture' .

Somalia highlights that the cross-border nature of AI's deployment and impact makes fragmented national approaches insufficient and potentially counterproductive. The speaker calls for unprecedented international cooperation as the only viable path to effective AI governance.

The speaker explicitly noted that AI's 'deployment and impact crosses national borders, requiring unprecedented international cooperation' . He also highlighted that for countries like Somalia, challenges including limited access to affordable computing infrastructure, unreliable energy, shortage of high-quality data, and limited digital and AI skills are 'challenges that no developing countries can overcome alone' .

India argues that the UN is uniquely positioned as the world's foremost platform for forging global consensus on AI governance, and that AI governance structures must reflect unity of purpose rather than divisions. The speaker warns that human divisions should not allow technology to outrun diplomacy.

The speaker stated that 'it is crucial that this global dialogue is taking place at the United Nations, which is the world's foremost platform for forging global consensus on issues of shared international concern' . He warned that 'AI governance structure must reflect unity of purpose and our divisions should not allow technology to outrun diplomacy' .

India's national strategy for AI and governance guidelines are built on the philosophy of inclusive growth for everyone without stifling innovation. The speaker emphasises that responsible AI governance must rest on strong foundational principles including human oversight, human rights, and prevention of misuse.

The speaker stated that 'India firmly believes that responsible AI governance must rest on strong foundational principles, preserving human oversight, upholding human rights, and preventing its misuse' . He noted that India's 'national strategy for AI and national AI governance guidelines embody the philosophy of AI for all, inclusive growth for everyone without stifling innovation' .

Ireland argues that the systems being developed today are shaping access to information, influencing economic opportunity, and affecting fundamental rights in ways that no single nation can address alone. A balanced, inclusive, and risk-based approach to AI governance is needed to harness AI's potential while managing its consequences.

The speaker stated that 'the governance of artificial intelligence cannot be addressed by any single nation alone' and that 'the systems being developed today at rapid pace are shaping access to information, influencing economic opportunity, and affecting fundamental rights' . She called for 'a balanced, inclusive, risk-based approach to AI governance' .

Ireland emphasises that a human-centric and human rights-based approach must be taken to AI development from the very beginning of the design process. This is reflected in Ireland's new national digital and AI strategy, which supports competitiveness and innovation while protecting data privacy and fundamental rights.

The speaker stated that 'it is vital that a human-centric and human rights-based approach is taken to the development of AI technology from the design stage onwards' . She noted that Ireland has developed 'a new national digital and AI strategy which supports competitiveness and innovation while also protecting data privacy and fundamental rights' .

Ireland sees the global dialogue as an opportunity to develop shared governance language that incorporates transparency, accountability, and oversight, and to advance multi-stakeholder cooperation towards safe and equitable access to AI technologies. The speaker also highlights the critical role of the Independent International Scientific Panel on AI.

The speaker stated that 'this dialogue can facilitate much-needed efforts towards shared governance language incorporating transparency, accountability and oversight' and that 'it can advance multi-stakeholder cooperation towards safe and equitable access to AI technologies' . She also welcomed the publication of the Independent International Scientific Panel on AI's first report and reiterated support for its vital work .

Bahrain, in collaboration with the Digital Cooperation Organisation, is leading exploratory work for a global AI treaty initiative that seeks to encourage trusted, inclusive, and interoperable approaches to AI safety. This initiative is guided by shared principles, multilateral collaboration, and respect for national security.

The speaker stated that 'in collaboration with the Digital Cooperation Organization, the GCO, Bahrain is leading the efforts of exploratory work for initiating a study for a global AI treaty initiative aimed at strengthening international dialogue and cooperation on the governance of AI-related risks' . He noted the initiative 'seeks to encourage trusted, inclusive, and interoperable approaches to AI safety guided by shared principles, multilateral collaboration, and respect for national security' .

Bahrain is developing its National Artificial Intelligence Strategy and has adopted the Gulf Cooperation Council AI Ethics Framework in 2025, incorporating AI and emerging technologies as strategic priorities within its National Telecommunications Plan. The speaker highlights Bahrain's strong digital foundations as enabling trusted AI adoption.

The speaker noted that 'Bahrain is currently developing its National Artificial Intelligence Strategy and adopted the Gulf Cooperation Council AI Ethics Framework in 2025' . He also stated that Bahrain's National Telecommunications Plan 'incorporates AI and emerging technologies as strategic priorities' , and highlighted achievements including ranking amongst the highest growth in the ITU ICT Development Index and achieving Tier 1 role modelling status in the Global Cybersecurity Index .

Mozambique calls for a global AI ecosystem founded on inclusive governance, capacity development, and trust as three essential pillars. The speaker emphasises that if access to AI remains concentrated among a few countries and corporations, it risks creating a technological divide that leaves developing countries further behind.

The speaker warned that 'if access to AI remains concentrated among a few countries and a handful of corporations, we risk creating a technological divide that will leave developing countries further behind' . He called for 'a global AI ecosystem founded on three essential pillars' including inclusive governance, capacity development, and trust .

Mozambique argues that trust in AI systems must be built on a foundation of human dignity, transparency, accountability, respect for human rights, cybersecurity, and international law. The speaker frames this as one of three essential pillars for a global AI ecosystem.

The speaker stated that 'trust' is one of the three essential pillars of a global AI ecosystem, and that 'AI must remain firmly anchored in human dignity, transparency, accountability, respect for human rights, cybersecurity, and international law' . He also noted that 'history will judge our generation not by how quickly we developed AI, but how wisely we governed it' .

Mozambique proposes a framework for a global AI ecosystem built on three essential pillars: inclusive governance that reflects all nations' voices, capacity development through investment in digital infrastructure and local talent, and trust anchored in human dignity and international law. The speaker argues that the decisions made today will determine whether AI narrows or widens inequalities.

The speaker outlined three essential pillars: first, 'inclusive governance' where 'global AI rules should reflect the voices and realities of all nations, including developing countries'; second, 'capacity development' requiring 'investment in digital infrastructure, affordable connectivity, reliable energy, quality education, digital literacy, research innovation, and development of local talent'; and third, 'trust' anchored in 'human dignity, transparency, accountability, respect for human rights, cybersecurity, and international law' .

Mozambique views AI as a strategic enabler for sustainable development across multiple sectors, including agriculture, healthcare, education, climate resilience, disaster readiness, and public administration. The country is developing a national AI strategy to guide this application.

The speaker stated that 'for Mozambique, AI is a strategic enabler for sustainable development' and that they are developing a national strategy of AI to apply it to 'strengthen agricultural productivity, improve healthcare delivery, enhance education, reinforce climate resilience, support disaster readiness, modernize public administration, and expand opportunities for millions of our citizens' .

Chad comes to the dialogue not as a spectator but as an actor with determination to lead Central Africa and the Sahel in the construction of global AI governance that is universal for all. The speaker emphasises that the future of AI must be constructed with the participation of all nations, in dignity, equity, inclusion, and sustainability.

The speaker stated that Chad comes 'not as a spectator here but as an actor and as an actor with determination in order to lead Central Africa and the Sahel in the construction of global governments of AI, which would be finally, at the end of the day, universal for all' .

Chad argues that ethics is not an obstacle to AI development but rather the foundation for its legitimacy, citing UNESCO's 2021 ethics recommendation as a key reference document for Africa. The speaker acknowledges that while this document is widely shared, unequal levels of development and the digital divide remain significant challenges.

The speaker referenced UNESCO's 2021 ethics recommendation, noting that 'Chad has subscribed to this document' and that 'ethics is not a barrier to development, rather, it's the foundation for its legitimacy' . He acknowledged that while this is 'a document of reference for the continent,' there are 'unequal levels of development' and a 'digital divide' that make the current situation 'insufficient' .

Chad acknowledges that AI is already transforming societies and economies with immense consequences across agriculture, health, governance, and public administration. However, the speaker argues that the concentration of AI development in the hands of very few violates human dignity and environmental sustainability.

The speaker stated that 'AI is already transforming our societies, our economies, and the freedom of the world, and its consequences are immense for agriculture, health, governance, public administration' . He warned that 'its development, of course, rests in the hands of very few, and this concentration is in violation of the dignity of human persons, sustainability for the environment' .

Chad sees the current situation as an opportunity for Central Africa to create regional governance frameworks that harmonise data and cybersecurity standards. The speaker notes that Chad is preparing a national AI and data governance strategy aligned with a federal-level AI strategy.

The speaker stated that 'Chad doesn't have a general framework' but that 'this is an opportunity, though, for Central Africa to create regional governance, to harmonize data and cybersecurity' . He noted that Chad is 'elaborating a national strategy for AI governance and data governance, which is aligned with our AI strategy at a federal level' and is 'working together to create regional governance' .

Thailand argues that the greatest risk to trust in AI is the fragmentation of different technologies, standards, and governance approaches, which risks fragmenting trust itself. To address this, Thailand proposes serving as a sandbox where global principles can be translated into practical implementations in real time.

The speaker stated that 'the greatest risk to that trust is fragmentation' of 'different technologies, different standards, especially different approaches to governance' and that 'if we allow AI governance to become fragmented, we risk fragmenting trust itself' . He extended 'an invitation to the United Nations and all member states' to 'use Thailand's current situation as a sandbox for AI governance' where 'global principles can be translated into practical implementations real-time' .

Thailand has taken concrete steps to build AI governance institutions and demonstrate regional leadership, including establishing an AI Governance Practice Centre, hosting the 2025 UNESCO Ethics of AI Forum, and chairing the ASEAN meeting that delivered the expanded Gen AI Guide. These efforts reflect Thailand's commitment to making governance a priority.

The speaker noted that Thailand 'established our AI governance practice center, AIGPC, hosted the 2025 UNESCO Ethics of AI Forum, and chaired the ASEAN meeting that delivered the expanded Gen AI Guide initiated by Singapore' . He also highlighted practical AI applications including using AI to address rainbombs, flash floods, wildfires, and scam operations .

Indonesia calls for a range of practical measures to enable developing countries to harness AI, including capacity building, technology partnerships, access to cloud and computing resources, local language AI development, and innovative voluntary financing mechanisms. The speaker emphasises that the principle of 'capacity before compliance' should guide global AI governance.

The speaker called for 'practical cooperation, capacity building, technology partnership, access to cloud and computing, resources, local language, AI development, and innovative voluntary financing mechanism' . She stated that 'for Indonesia the principle is clear, capacity before compliance' .

Indonesia argues that trustworthy AI governance must respect different national contexts and levels of development, making a one-size-fits-all approach neither realistic nor equitable. The speaker calls for interoperability rather than uniformity as the guiding principle for global AI governance.

The speaker stated that 'Indonesia underlines that trustworthy AI requires interoperability, not uniformity' and that 'global AI governance must respect different national contexts and levels of development' . She explicitly stated that 'a one-size-fits-all approach is neither realistic nor equitable' .

Indonesia is developing presidential regulations on a national AI roadmap and ethics to provide strategic direction for AI development across all sectors. The speaker also highlights Indonesia's commitment to child protection in the digital space as a concrete example of responsible AI governance.

The speaker stated that 'Indonesia is developing the presidential regulations on its national AI roadmap and ethics to provide strategic direction of AI development across all sectors and to promote responsible, ethical, trustworthy, and human-centric AI' . She also noted that President Prabowo enacted government regulations on electronic system governance in child protection, with '5 million accounts reportedly have been deactivated by platforms within the last five months' .

Rwanda highlights the stark concentration of global AI infrastructure, with over half of the world's data centres located in a handful of wealthy countries. The speaker argues that Africa must look inward for solutions and adapt, rather than relying solely on external resources.

The speaker stated that 'over half of the world's data centers are concentrated in a handful of wealthy countries' and that 'for us in Africa, it is in our interest to look inwards for solutions and adapt' .

Rwanda has put in place a comprehensive regulatory framework for AI, including a national AI policy, data protection and privacy law, and a newly established national AI agency. The speaker also emphasises the importance of training people to remain competitive in the AI era.

The speaker stated that 'in Rwanda, we've put in place strong regulatory frameworks, starting with our national AI policy and data protection and privacy law' . She noted that 'with our new established national AI agency, we intend to accelerate the deployment of this technology across multiple sectors' and that 'we are also training our people to remain competitive' .

Rwanda argues that regulatory fragmentation slows Africa down, making it difficult for entrepreneurs, data scientists, and engineers to break through borders when every market has a different rule book. To address this, 49 African countries adopted the Africa Declaration on Artificial Intelligence in Kigali to harmonise regulations across the continent.

The speaker stated that 'it is hard to break through borders when every market has a different rule book' and that 'fragmentation slows us down' . She noted that 'last year in Kigali, 49 African countries, together with the African Union and the Smart Africa' adopted 'the Africa Declaration on Artificial Intelligence' .

Iraq argues that safe and ethical AI governance must rest on three main pillars: ensuring AI is not used for lethal weapons or warfare, protecting families and children from harmful algorithms, and investing in human capacity so that humans are the first to fully benefit from AI technologies.

The speaker outlined three pillars: first, 'we need to be far away from warfare technologies' and 'make sure that AI is only a tool for life, not a tool for developing lethal weapons'; second, 'protect families and children' by ensuring 'algorithms are safe tools which preserve the innocence of childhood'; and third, 'invest in the development of human capacity to make sure that humans are the first to fully benefit from these new technologies' .

Pakistan argues that inclusion in AI governance without concrete investment is meaningless, calling for resource mechanisms, access to open source models, shared compute initiatives, support for data infrastructure, regulatory sandboxes, and South-South cooperation. These should be anchored in the Sustainable Development Goals.

The speaker stated that 'inclusion without investment is an empty promise' and called for 'concrete resource mechanisms, access to open source models, shared compute initiatives, support for data infrastructure and national AI strategies, regulatory sandboxes, fit for local conditions, and South-South and South-East-West cooperation' . She noted these should be 'anchored explicitly in the sustainable development goals' .

Côte d'Ivoire highlights AI's enormous economic potential, citing a study suggesting it could contribute $15.7 trillion to the global economy by 2030, but argues that this potential must be distributed equitably. The speaker insists that AI must not be a privilege for some economies and platforms but a lever for shared productivity.

The speaker cited 'a recent study' suggesting 'AI could contribute some $15.7 trillion to the economy by 2030' . He stated that 'AI must not be the privilege for some economies and platforms' but 'must be a lever for shared productivity' .

Côte d'Ivoire argues that AI must be accessible to all economies and serve as a lever for shared productivity and development, not concentrated in the hands of a few. The speaker outlines Côte d'Ivoire's ambition to use AI to transform its economy and reduce poverty.

The speaker stated that 'AI must not be the privilege for some economies and platforms' and 'must be a lever for shared productivity' . He noted that Côte d'Ivoire 'has the ambition to become an emerging country with a significant drop in poverty' and hopes 'that artificial intelligence will contribute to the deep-rooted transformation of our economy' .

Senegal speaks on behalf of a broad Global Network for cooperation on AI capacity building, which includes partners from Brazil, Cambodia, China, Ethiopia, Germany, and many other countries. The speaker welcomes the emphasis on capacity building as essential for reducing the AI and digital divide.

The speaker stated that Senegal is expressing itself 'on behalf of the Global Network of cooperation for the strengthening of AI capacity' with centres involving partners including Brazil, Cambodia, China, Ethiopia, Germany, Guinea, India, Japan, Kenya, Rwanda, Niger, Saudi Arabia, Vietnam, and the EU . He welcomed 'the accent that has been made on capacity building for AI, which is an essential factor in reducing the AI and digital divide' .

Senegal calls on a wide range of stakeholders, including member states, academia, philanthropic organisations, and the private sector, to actively support the AI capacity building network through resources, expertise, and centres of excellence. The speaker also thanks the UN for its support in making AI a lever for cooperation and development.

The speaker invited 'member states, as well as academia, philanthropic organizations, the private sector, and all partners to actively support our network through resources, expertise, and centers of excellence, as well as partnerships' . He also thanked 'the Emerging Digital Technology UN Office for its support and making AI a lever for cooperation and development' .

The Maldives argues that for small island developing states, AI must be built on a foundation of digital public infrastructure, trusted data, cybersecurity, and clear accountability, or it will deepen inequality rather than reduce it. The speaker emphasises that inclusion begins with these foundations.

The speaker stated that under the Maldives 2.0 roadmap, 'AI is treated as one layer within a broader transformation' that 'must rest on digital public infrastructure, trusted data, cybersecurity, and clear accountability' . He warned that 'without AI deepens inequality rather than reducing it, concentrating benefits in already capable institutions and leaving others further behind' .

The Maldives argues that interoperability should be treated as a governance principle, not merely a technical standard, as it allows countries to build AI capabilities incrementally and adapt new tools on a stable, sovereign foundation. This is one of three priorities the speaker emphasises from the Maldives' experience.

The speaker stated that countries should 'treat interoperability as a governance principle' as 'it lets countries build incrementally and adapt new tools on a stable, sovereign foundation' . He also emphasised the need to 'invest in foundations, not only applications' and to 'treat capacity building as a serious long-term concern' .

The Maldives has taken concrete steps to build AI governance capacity, completing South Asia's first national AI readiness assessment with UNESCO and establishing the Maldives AI Lab to move from fragmented experimentation towards coordinated, governed capability across government. The speaker frames this as grounding ambition in action.

The speaker stated that 'with UNESCO, we completed South Asia's first national AI readiness assessment' and that the Maldives 'established our national AI and data competency center, the Maldives AI Lab, to move from fragmented experimentation towards coordinated, governed capability across government' .

Chile views AI as a form of increased intelligence designed to optimise human capacity, resources, and generate strategic economic value within a democratic governance framework. The speaker highlights Chile's strong enabling infrastructure, including 95% connectivity, 65% renewables in the electricity grid, critical minerals, and a growing startup ecosystem.

The speaker described AI as 'increased intelligence at all designed to optimise human capacity, optimise resources and generate strategic economic value' . She highlighted that Chile has '95% connectivity across all of our territory' and '65% renewables in our electricity grid,' along with critical minerals, a consolidated scientific ecosystem, and growing startups .

Chile proposes a set of concrete governance tools for the international community, including multilateral sandboxes for AI testing under common rules, open data ecosystems, ongoing education with emphasis on technical education, and identification of serious standards and certifications. The speaker argues that AI can distribute opportunities rather than concentrate power.

The speaker proposed 'seeing education as an ongoing process, putting emphasis on technical education, identifying serious standards and certifications such as ISO, AI Verify and others,' 'to establish a multilateral network of sandboxes for AI testing under common rules which would facilitate experiments,' and 'promoting open data ecosystems' . She stated that 'AI alone will not create divides' and 'can concentrate power or it can distribute opportunities' .

Zimbabwe strongly supports the Secretary General's call for a global fund for AI to build skills, data, and innovation, and calls for three specific actions: a global AI capacity building fund, a mutual recognition framework for AI ethics standards, and an interoperability by default pledge. The speaker frames governance as the guarantor of trust, not the enemy of innovation.

The speaker stated that Zimbabwe 'strongly supports the Secretary General's call for a global fund for AI to build skills, data, and innovation' . She called for three actions: 'an AI global, AI capacity building fund,' 'a mutual recognition framework for AI ethics standards,' and 'an interoperability by default pledge' .

Zimbabwe launched its National Artificial Intelligence Strategy in March, providing a comprehensive framework to harness AI for inclusive growth and digital sovereignty. The strategy is supported by dedicated startup funds, innovation initiatives, and special economic zones designed to attract both local startups and international AI players.

The speaker stated that 'in March this year, Zimbabwe launched its National Artificial Intelligence Strategy, a comprehensive framework to harness AI for inclusive growth and digital sovereignty' . She noted that Zimbabwe is 'rolling out dedicated startup funds and innovation initiatives' and has 'established special economic zones geared towards AI-inspired investment' .

Zimbabwe is actively investing in youth entrepreneurship in the AI space through dedicated startup funds, innovation initiatives, and special economic zones. These measures are designed to give young entrepreneurs and tech developers the capital, mentorship, and incubation support they need to bring their ideas to market.

The speaker stated that Zimbabwe is 'rolling out dedicated startup funds and innovation initiatives that will give young entrepreneurs and tech developers the capital, mentorship, and incubation support they need to bring their ideas to market' . She also noted that the government has 'established special economic zones geared towards AI-inspired investment, begged by the incentives and infrastructure required to draw both local startups and international AI players to our shores' .

The United States argues that voluntary cooperation between government and industry is the only AI security approach agile enough to meet emerging challenges without stifling innovation, because industry is intimately aware of the state, scope, and trajectory of AI developments in ways that government is not.

The speaker stated that 'industry is intimately aware of the state, scope, and trajectory of AI developments' while 'government is not,' and that 'voluntary cooperation between the two is the only AI security approach agile enough to meet challenges as they arise without stifling innovation' .

The United States emphasises its commitment to a pro-innovation environment that allows AI companies to meet people's real needs, combined with a national policy framework that protects children, prevents censorship, and supports workers and communities. The Trump administration is also passionate about sharing the benefits of American AI globally.

The speaker stated that 'the Trump administration is working with American legislators as they construct a national policy framework for AI that protects children, prevents censorship, respects information, and promotes health and well-being' . She noted that 'the Trump administration is passionate about sharing the benefits of American AI with the rest of the world' and that 'shared efforts to advance the diffusion of the full AI stack across partner countries will become an occasion for bilateral diplomacy, international development, and global economic dynamism' .

Singapore has taken concrete steps to build trust in AI by publishing guidelines on the responsible use of data for AI models and on AI agents, and by hosting the International Scientific Exchange for AI Safety Experts. This exchange brought together experts from academia, industry, and civic organisations to develop the Singapore Consensus on AI Safety Research Priorities.

The speaker stated that 'Singapore has published guidelines on the responsible use of data for AI models and on AI agents' and that Singapore 'partner[s] international bodies like the ISO and IEC to standardise the methodology for testing generative AI systems' . She noted that Singapore 'hosted the International Scientific Exchange for AI Safety Experts from academia, industry and civic organisations who have in turn developed the Singapore Consensus for AI Safety Research Priorities' .

Italy argues that the challenge of AI governance requires principles that are truly universal and shared by all nations, not imposed by a few powerful actors. The speaker calls for treating AI governance as a shared responsibility rather than something to watch unfold.

The speaker stated that 'a challenge of this magnitude requires principles that are truly universal and shared by all' and 'not imposed by the few, and that every country can then adapt to its own governance experience' . He argued that 'it is time we treated it as a responsibility to share' .

Italy's approach to AI governance centres on putting the person at the centre, designing AI with a security by design approach to prevent malicious uses, and involving all stakeholders including governments, businesses, civil society, and the research community. The speaker also emphasises the importance of closing the digital divide without opening a new AI divide.

The speaker stated that 'for Italy, governance of this kind means, first and foremost, putting the person at the center' and 'designing artificial intelligence according to a security by design approach, because trust in these systems is built by preventing malicious uses, not by chasing after them' . He called for 'involving all stakeholders, governments, businesses, civil society, the research community, because no one can write these rules alone' .

Italy has enacted a national AI law that aligns its legal framework with the European AI Act and entrusts governance of the sector to two existing agencies. The speaker frames rules not as brakes on innovation but as guardrails that make innovation safer and faster.

The speaker stated that 'our law on artificial intelligence aligns our national legal framework with the European AI Act and entrusts governance of the sector to two existing agencies' . He argued that 'the rules we build together are not a brake on innovation' but 'the guardrail that makes it safer and faster' .

Mali argues that AI is not neutral and reflects the biases of those who conduct the training, meaning AI systems based on non-representative data can amplify and worsen discrimination. The speaker frames data sovereignty not as protectionism but as a matter of dignity and sovereignty for nations of the Global South.

The speaker stated that 'AI is not neutral' and 'actually reflects the biases of those who are doing the training,' meaning 'AI systems, which are based on non-representative data, can amplify discrimination' and even 'worsen discrimination' . He argued that 'the south countries, countries of the global south, must have a sovereign right to collect, store, and analyze their data' and that 'it's not just a protectionist aspect' but 'a matter of the dignity and the sovereignty of each nation' .

Mali calls for the adoption of an international framework for AI governance with clear rules that respect the rights of developing countries and their sovereign rights over data. The speaker also calls for the representation of African languages and cultures in AI systems and a dedicated development space for least developed countries.

The speaker called for 'the adoption of an international framework for AI governance, which respects all and which has clear international rules respecting the rights of developing countries and the sovereign rights of countries to equitably be represented in their data' . He also called for 'a dedicated space in the development area which is focused on least developed countries' and for 'represent[ing] African languages and cultures in AI' .

The Netherlands argues that interoperability should be the organising principle for the next phase of AI governance, meaning frameworks built in different jurisdictions on common international standards that can recognise and work with each other. This is distinct from harmonising away legitimate differences or imposing one rulebook on the world.

The speaker stated that 'interoperability' is 'the organizing principle for this next phase' and that 'it does not mean one rulebook for the world, nor harmonizing away legitimate differences' but rather 'frameworks built in different jurisdictions, built on the common international standards to rights, safety, transparency, and accountability' that 'can recognize and work together with each other' .

The Netherlands argues that AI governance requires genuine multi-stakeholder participation, with civil society, researchers, and the technical community having real seats at the table, not just token inclusion. The speaker emphasises that these actors bring knowledge that governments lack.

The speaker stated that 'AI governance cannot be written by governments or technology companies alone' and that 'civil society, researchers and the technical community bring knowledge that we lack and it's our task to give this multi-stakeholder real seats at the real table' .

The Netherlands outlines a comprehensive set of principles for responsible AI governance, including being responsible, safe, secure, trustworthy, human-centric, reliable, privacy-preserving, and accountable. The speaker notes that the international community has already constructed building blocks for global AI governance through the OECD, G20, and UNESCO.

The speaker stated that the Netherlands' approach is 'informed by the conviction that AI should be responsible, safe, secure and trustworthy, human-centric, reliable, privacy-preserving and accountable' . She noted that 'the international community has constructed the building blocks of a global AI governance through, for example, the OECD and the G20 AI principles and the UNESCO recommendation' .

The Netherlands has implemented a public algorithm register as a concrete transparency measure, documenting around 1,500 algorithms across nearly 350 public sector bodies. The speaker offers this as an example that others can adapt, not as a definitive answer.

The speaker stated that 'our public algorithm register now documents around 1,500 algorithms across nearly 350 public sector bodies' . She offered this as 'an example we offer not as an answer but as something someone can adapt' .

Malawi faces a profound dilemma in approaching AI governance, with compounded barriers including persistently low foreign direct investment, unreliable energy supply, inadequate connectivity, prohibitive costs for data and devices, and limited digital literacy. The speaker argues that these barriers widen the digital gap even before AI adoption begins.

The speaker stated that Malawi's 'challenges are compounded by persistently low foreign direct investment and reliable energy supply, inadequate connectivity, prohibitive cost for data and devices, and limited digital literacy among our people' . He noted that 'these barriers widen the digital gap even before we ever begin' and that Malawi faces 'an impossible choice, accept unacceptable risks or be left behind' .

Namibia argues that without inclusive digital foundations, AI risks becoming a tool for the few rather than a catalyst for shared prosperity. The speaker calls for global cooperation to expand open data sets, support shared AI infrastructure, and promote responsible open source AI models.

The speaker stated that 'without inclusive digital foundations, AI risks becoming a tool for the few, not a catalyst for shared prosperity' . She called for 'global cooperation to expand open data sets for development, support shared AI infrastructure, including regional AI hubs and maybe even AI capacity fund for developing states and promote responsible open source AI models that empower innovation without compromising safety' .

Namibia raises the specific concern that bias in AI systems is threatening gains in women's empowerment, with increasing digital violence and concerns around technology-facilitated or AI-facilitated gender-based violence. The speaker calls for both public and private sector collaboration to address these risks.

The speaker stated that 'bias in AI systems is threatening these gains due to an increase in digital violence with great concerns around technology-facilitated or AI-facilitated gender-based violence' . She called for 'both public and private sector collaboration to build a future with AI capabilities that will perform the mundane tasks in order to free up humans for innovation, for empathy, and creativity' .

Uganda calls for concrete partnership and cooperation on compute access, capacity building, and financing for development from all countries present at the dialogue. The speaker highlights that Africa holds less than 1% of the world's AI compute, making Uganda's 2040 economic vision extremely difficult to achieve without external support.

The speaker stated that 'Africa holds less than 1% of the world's AI compute and our 2040 goal cannot be achieved easily on the less than 1% of the world's capability' . He called for 'concrete partnership cooperation on input access, capacity building, and financing for development from all countries who are present today' .

Uganda highlights the stark reality that Africa holds less than 1% of the world's AI compute, making it extremely difficult for countries like Uganda to achieve their ambitious economic visions through AI. The speaker calls for concrete partnerships on compute access, capacity building, and financing to bridge this gap.

The speaker stated that 'Africa holds less than 1% of the world's AI compute and our 2040 goal cannot be achieved easily on the less than 1% of the world's capability' . He noted Uganda's ambition 'to take our economy to the future $500 billion through support of AI' .

Uganda favours interoperable cooperation and standards-based approaches over fragmented national approaches for achieving safe, secure, and trustworthy AI. The speaker aligns Uganda with the African group's statement and emphasises the importance of bridging AI divides in compute, data, and skills.

The speaker stated that 'on safe, secure, and trustworthy AI, we favor inter-portal cooperation, standards based on approaches over fragmented ones' . He also noted that Uganda 'focuses on bridging the divides those are the AI divides coming from compute, data and skills' .

The United Kingdom argues that the UN has an indispensable role to play in AI governance through three specific functions: building a common evidence base through the Independent Scientific Panel on AI, maintaining a strong foundation of principles for trustworthy AI, and helping countries access the support, evidence, and expertise they need to benefit from AI.

The speaker outlined three roles for the UN: 'helping build a common evidence base' through the International Independent Scientific Panel on AI; 'helping maintain a strong foundation of the principles for trustworthy AI' by 'connecting and consistently operationalising these and translating them to reality in all countries'; and 'helping countries access the support, evidence and expertise they need to benefit from AI' .

The United Kingdom emphasises the critical role of the Independent International Scientific Panel on AI in helping all countries develop a shared understanding of AI's capabilities and risks. The speaker also highlights the UK's AI for Development programme as a concrete example of partnership-based AI governance.

The speaker stated that 'the International Independent Scientific Panel on AI can help ensure all countries have a shared understanding of AI's capabilities and risks' . She also highlighted the UK's collaboration 'with some of the brightest minds across Africa and Asia through our AI for Development programme,' including 'helping build AI tools in more than 40 African languages, supporting 13 AI labs' .

As co-chair, Rein Tammsaar articulates the overarching purpose of the dialogue as a platform for sharing diverse national and regional perspectives on AI governance and fostering expectations for international cooperation. This framing establishes the dialogue as a space for collective learning and exchange in a rapidly evolving field.

The co-chair stated that the purpose of the session is 'to share national and regional perspectives, priorities, experiences, and lessons learned related to AI governance, while also reflecting on the expectations for international cooperation in this rapidly evolving field' .

At the close of the high-level governmental plenary segment, Rein Tammsaar, on behalf of the co-chairs, acknowledged and appreciated the breadth of contributions from member states covering AI governance, international cooperation, and the opportunities AI presents. This closing remark underscores the value placed on inclusive participation in the dialogue.

The co-chair stated that 'on behalf of the co-chairs, we would like to express our sincere appreciation to all member states for their contributions and for sharing national, regional, and collective perspectives on AI governance, international cooperation, and the opportunities' .

As co-chair, Egriselda López systematically gave the floor to ministers and high-level representatives from a wide range of countries, including from Africa, Asia, Europe, and the Americas, ensuring that diverse regional voices were heard in the dialogue. This procedural role reflects the commitment to inclusive participation that underpins the global dialogue's mandate.

The co-chair introduced and gave the floor to representatives from the United States, Singapore, Ireland, Zimbabwe, Uganda, the United Kingdom, Mozambique, Malawi, Mali, Italy, the Netherlands, and Somalia, among others, consistently acknowledging each speaker before passing to the next .