University of Huddersfield, United Kingdom

A successful first Global Dialogue on AI Governance should deliver more than broad agreement on principles. It should produce a credible foundation for inclusive, practical, and accountable global cooperation. First, success would mean meaningful Global South participation, not just symbolic representation. Countries like Nigeria and others in Africa, Asia, Latin America, and small island states should help shape priorities, not simply react to frameworks developed elsewhere. Second, the Dialogue should identify clear shared priorities for action. These should include capacity-building, access to digital infrastructure and compute, support for local research and innovation, multilingual and culturally inclusive AI, and safeguards for high-impact uses of AI in areas such as public services, education, healthcare, finance, and security. Third, success would require agreement that AI governance must include accountability mechanisms. This means promoting transparency, human oversight, impact assessments, and access to remedy where AI systems affect people's rights, opportunities, or wellbeing. Fourth, the Dialogue should recognize that one-size-fits-all governance will not work. It should support context-sensitive approaches while still upholding common global values such as fairness, safety, dignity, and human rights. Finally, the first Dialogue would be successful if it creates momentum for sustained cooperation: a clear roadmap, follow-up actions, and continued engagement with governments, civil society, academia, youth, and underrepresented regions. In short, success would mean moving from conversation to commitment, and from inclusion in language to inclusion in practice.

• Social, economic, ethical, cultural, linguistic and technical implications of AI
• AI capacity-building
• Protection and promotion of human rights
• Transparency, accountability, and human oversight

AI capacity-building, social, economic, ethical, cultural, linguistic and technical implications of AI, protection and promotion of human rights, and transparency, accountability, and human oversight reflect the most urgent priorities for Nigeria and many other Global South countries. First, AI capacity-building is essential because effective participation in AI governance depends on more than policy ambition. It requires skills, compute, infrastructure, data systems, and institutional readiness. The World Bank has emphasized that inclusive AI ecosystems depend on connectivity, compute, context, and competency, all of which remain unevenly distributed across developing countries. Second, I chose the area on social, economic, ethical, cultural, linguistic and technical implications because AI systems often reflect unequal global realities. For many countries in Africa and the wider Global South, concerns include bias, exclusion, labour-market disruption, weak local relevance, and the underrepresentation of local languages and cultural contexts. UNESCO's multilingualism work highlights the importance of equitable digital representation and support for diverse languages and communities. Third, human rights must remain central, especially where AI is being used in public services, education, finance, security, and social protection. The UN has framed the Global Dialogue around human rights, safety, and ensuring that AI benefits all. Finally, transparency, accountability, and human oversight are critical because people must be able to understand, question, and seek remedy for high-impact AI decisions. For my context, these priorities best combine development needs, ethics, rights protection, and practical governance safeguards.

Yes. In my view, at least three cross-cutting issues deserve more explicit attention. 1. Concentration of compute, cloud infrastructure, and market power: This is not fully captured by the listed themes, yet it shapes almost every other issue. The World Bank has highlighted a severe global divide in compute capacity, and its work on digital public infrastructure warns that vendor lock-in can lead to high costs, limited interoperability, and reduced policy flexibility. For many Global South countries, this is a question of digital sovereignty as much as innovation. 2. Environmental sustainability and resource use: AI governance should explicitly address energy use, water consumption, extractive supply chains, and e-waste. UNESCO has emphasized that generative AI is increasingly resource-intensive and notes that more efficient approaches can dramatically reduce energy use, in some cases by up to 90%. For developing countries facing infrastructure and climate pressures, sustainable AI should be treated as a core governance issue, not a side concern. 3. Public procurement and implementation governance: Many governments will experience AI primarily through procured systems rather than systems they build themselves. The UN Human Rights Council has noted that the question is not whether to regulate AI procurement and deployment, but how to do so consistently with international human rights law. This is especially important in sectors such as welfare, education, health, and security. Taken together, these issues affect whether AI governance is genuinely inclusive, affordable, sustainable, and workable for the Global South.

In Nigeria, and more broadly across the Global South, the main governance gap is the mismatch between the speed of AI adoption and the strength of the foundations needed to govern it well. AI interest is growing in public services, education, finance, agriculture, and health, and Nigeria has taken important steps through its National AI Strategy and the Nigeria Data Protection Commission's 2025 GAID. These are significant opportunities because they create a basis for more responsible, rights-aware, and development-oriented AI adoption. The most significant challenge, however, is capacity. The World Bank's 2025 AI Foundations report shows that inclusive AI depends on connectivity, compute, context, and competency. These remain unevenly distributed across many developing countries, including in Africa. In practice, this means limited infrastructure, unreliable power, high costs of cloud and compute, shortages of advanced AI skills, and weak institutional capacity to evaluate or audit imported systems. As a result, countries risk becoming users of AI without having equal influence over how it is designed, deployed, or governed. A second major challenge is inclusion and rights protection. Many local languages and cultural contexts remain underrepresented in AI systems, which can reduce accuracy, fairness, and accessibility. UNESCO's multilingualism roadmap highlights the need for language community involvement, data sovereignty, and capacity-building for under-resourced languages. For Nigeria, this is highly relevant given its linguistic diversity. It creates both a risk of exclusion and an opportunity to build more locally grounded, culturally relevant AI systems. Overall, the biggest opportunity is to build AI governance that is not only innovative, but inclusive, accountable, and rooted in local realities. The biggest risk is adopting AI faster than we build the institutions, safeguards, and public capacity needed to govern it well.

The AI Dialogue can play a valuable role by becoming a bridge-building mechanism; one that connects technical debate, policy discussion, and development realities across regions. The UN describes the Dialogue as an inclusive platform for governments and stakeholders to discuss international cooperation on AI governance, following General Assembly resolution 79/325. That gives it legitimacy and convening power that few other forums have. From my perspective, its most important contribution would be to help develop shared understanding across different governance approaches while ensuring that countries in the Global South are not only consulted, but heard in shaping priorities. The Dialogue can create space for countries with different levels of capacity, regulatory maturity, and technological development to identify common concerns such as safety, human rights, accountability, and access to opportunity. It can also advance cooperation by focusing attention on practical enablers of inclusion, especially capacity-building, access to infrastructure and compute, and support for locally relevant innovation. This matters because the benefits of AI remain highly concentrated, while many developing countries risk exclusion from shaping and benefiting from the technology. UNCTAD has warned that AI's gains are unevenly distributed, and the UN has linked AI governance to the need to narrow widening digital divides. Finally, the Dialogue will be most useful if it produces continuity and follow-through: clearer priority areas, stronger links to the Independent International Scientific Panel on AI, and sustained engagement beyond one meeting. Success in international cooperation will come not just from dialogue itself, but from building a trusted pathway from discussion to coordination, capacity support, and concrete action.

The AI Dialogue should build on existing initiatives rather than start from zero. In particular, it should connect with the UNESCO Recommendation on the Ethics of AI, which already provides a global normative baseline on human rights, transparency, fairness, and human oversight across all 194 UNESCO Member States. It should also link closely to the Global Digital Compact follow-up architecture, especially the Independent International Scientific Panel on AI, which was established alongside the Dialogue by General Assembly resolution 79/325. It should also build on the OECD AI Principles and the integrated GPAI/OECD.AI ecosystem, which provide practical guidance, policy tracking, and work on interoperability across jurisdictions. In addition, the Dialogue should connect with ITU's AI for Good platform and related standards work, which already convene governments, technical experts, UN agencies, and industry around implementation, skills, and standards. From an African and Global South perspective, the Dialogue should also engage regional frameworks such as the African Union Continental AI Strategy, which reflects Africa-specific priorities around inclusion, development, ethics, and context-sensitive governance. This would help ensure that global cooperation is informed by regional realities rather than dominated by a few advanced economies. The added value of the AI Dialogue would be its UN legitimacy, inclusiveness, and convening power. It can connect these fragmented initiatives, amplify underrepresented voices, and translate existing principles into a more coherent global cooperation agenda. Its unique role should be to bridge norms, evidence, regional experience, and capacity-building so that AI governance becomes more coordinated, development-oriented, and globally representative.

Different stakeholders should contribute in ways that reflect their strengths but also ensure balanced influence. Governments should bring policy experience, regulatory priorities, and public-interest concerns; the private sector should provide evidence on deployment, risk management, and transparency practices; academia and the technical community should contribute independent research, standards expertise, and evaluation methods; civil society, labour groups, and human rights advocates should surface real-world impacts on affected communities; and youth, indigenous peoples, persons with disabilities, and stakeholders from developing countries should be involved early enough to shape priorities, not simply respond to them. UNESCO's AI ethics framework emphasizes human rights, diversity, fairness, and human oversight, while the IGF's capacity-development work highlights the importance of enabling participation from developing countries and underrepresented groups. In terms of format, the AI Dialogue should combine a high-level political segment with smaller, interactive multistakeholder roundtables. The UN concept note already points toward a plenary meeting, a governmental segment, thematic discussions, and multistakeholder consultation; this is a good foundation. To make it more effective, the process should include regional pre-dialogue consultations, open written submissions, strong hybrid participation, multilingual interpretation, accessible documentation, and financial support or fellowship-style mechanisms for participants from the Global South. The IGF fellowship model and NETmundial+10's São Paulo Multistakeholder Guidelines offer useful lessons on inclusion, trust-building, and meaningful participation. Finally, the Dialogue should produce a short public outcome document that clearly identifies areas of convergence, areas of disagreement, and practical next steps. That would help turn participation into continuity and cooperation rather than a one-off event.

Yes. The most underrepresented voices are still those from developing countries, especially African countries, least developed countries, small island developing States, and communities that experience AI mainly as users of imported systems rather than as designers of them. In practice, this also includes local civil society groups, public-interest researchers, small innovators, and policymakers with limited technical capacity. Countries do not engage with AI from equal positions, and African priorities are often rooted in infrastructure, data, and institutional gaps rather than only frontier-model risks. Also underrepresented are speakers of African, Indigenous, and other low-resource languages, as well as Indigenous communities more broadly. UNESCO's multilingualism work explicitly stresses that underrepresented and Indigenous language communities must be able to access, benefit from, and fully participate in the digital world. Current UNESCO monitoring also shows that language representation in AI training resources remains limited across many countries. A further gap concerns persons with disabilities, women and girls in the Global South, older persons, youth outside major innovation hubs, and workers in informal or vulnerable sectors whose livelihoods may be reshaped by AI. UNESCO's accessibility and inclusion work emphasizes that participation from underrepresented groups, particularly women and girls in the Global South and people across age groups, is necessary if AI is to benefit all rather than widen existing divides. They could be included through funded participation, regional pre-dialogue consultations, multilingual interpretation, accessible formats, hybrid participation, and dedicated seats in thematic roundtables. Just as importantly, the Dialogue should treat community evidence as expertise: affected communities should not only be invited to speak, but should help define priorities, risks, and success measures from the outset.

Innovative formats should make the AI Dialogue more participatory, problem-solving, and evidence-based. The UN concept note already envisages a plenary, a high-level governmental segment, thematic discussions, and a multistakeholder consultation. Building on that, I would add scenario-based policy labs where mixed groups work through concrete use cases such as AI in education, health, welfare, elections, or low-resource languages, and then produce short practical recommendations. A second useful format would be moderated fishbowl dialogues and rotating roundtables. These would allow governments, private sector actors, academics, civil society, and affected communities to respond to one another directly instead of delivering isolated statements. NETmundial+10 emphasizes the need to improve mechanisms for consensus-building and to ensure that communities' voices meaningfully influence multilateral decision-making. A third format should be regional hubs with strong hybrid participation. Regional or national gatherings could feed live input into the main Dialogue, especially from Africa, Latin America, Asia-Pacific, and small island states. This would help address participation barriers linked to cost, visas, and time zones. IGF materials note that its fellowship programme is meant to build capacity in developing countries, and IGF stocktaking highlights that hybrid formats broaden inclusivity and diversity. Finally, I would recommend lightning evidence sessions and a live synthesis board. Short 2–3 minute interventions from youth, disability advocates, labour groups, local innovators, and low-resource language communities could be captured in real time under headings such as convergence, disagreement, and next steps. This would keep the process dynamic while ensuring that participation leads to a visible and actionable public record.

Examples already exist across different layers of AI governance. First, the UNESCO Recommendation on the Ethics of AI is a strong global normative example. It provides a shared baseline built around human rights, dignity, transparency, fairness, and human oversight, and it applies across UNESCO's 194 Member States. It is useful because it gives countries a common ethical reference point even when their legal systems differ. Second, the EU AI Act offers a concrete regulatory model. Its risk-based approach, phased application timeline, rules for general-purpose AI, and restrictions on prohibited practices show how broad principles can be translated into enforceable obligations. This is helpful for countries looking for practical legislative approaches to high-risk AI. Third, the NIST AI Risk Management Framework is a valuable practical tool. Because it is voluntary and focused on managing risks to individuals, organisations, and society across the design, development, use, and evaluation of AI systems, it offers an adaptable approach for both public and private sector actors. Fourth, from a Nigerian perspective, the Nigeria Data Protection Commission's GAID 2025 and related privacy-by-design work show how data governance can provide a foundation for AI governance. GAID emphasizes fairness, lawfulness, transparency, accountability, security risk assessment, and the use of Data Privacy Impact Assessments, which are all highly relevant for AI deployment. Finally, platforms such as OECD.AI and ITU's AI for Good offer practical added value by sharing policy evidence, standards, implementation practices, and partnerships. These kinds of knowledge-sharing mechanisms help countries learn from one another and avoid duplicating effort.