Binus School, Jakarta, Indonesia

A successful first Global Dialogue on AI Governance would, in my view, be one that moves beyond broad statements and produces concrete foundations for sustained international cooperation. The meeting should help clarify not only what responsible AI governance means in principle, but also how countries and institutions can work together despite differences in legal systems, technical capacity, and political context. One important outcome would be the identification of a shared set of governance priorities. These could include human rights protection, transparency, accountability, safety, data responsibility, and meaningful oversight of high-impact AI systems. The goal does not need to be complete uniformity, but rather a common language that helps governments, civil society, academia, and industry communicate more effectively and coordinate where possible. A second sign of success would be the collection of practical lessons and best practices. Many actors are already experimenting with regulatory frameworks, public-sector guidance, risk-based approaches, and technical standards. The Dialogue should help synthesize these experiences into examples that others can adapt, especially countries that are still building their own AI governance capacity. Inclusion would also be essential. The first Dialogue would only be truly successful if it amplifies perspectives that are often underrepresented in global technology discussions, particularly from the Global South, smaller and lower-capacity states, youth, educators, workers, and communities most likely to be affected by AI deployment. Governance discussions gain legitimacy when they reflect real social diversity rather than only the priorities of major powers and large technology actors. Finally, success would require continuity. The Dialogue should not end as a one-time conversation, but should create a pathway for follow-up, knowledge-sharing, and collaborative action. If participants leave with stronger trust, clearer priorities, and a practical agenda for continued engagement, then the first Global Dialogue on AI Governance will have achieved something meaningful and lasting.

• Safe, secure and trustworthy AI
• AI capacity-building
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Open-source software, open data and open AI models

I selected these themes because they address both the immediate risks of AI and the longer-term conditions needed for more equitable and effective global governance. The UN's framework for the Dialogue explicitly includes safe, secure and trustworthy AI; AI capacity-building; the social, economic, ethical, cultural, linguistic and technical implications of AI; and open-source software, open data and open AI models as core discussion areas. First, safe, secure and trustworthy AI is essential because public trust will shape whether AI is accepted, regulated responsibly, and used for social benefit. Governance discussions must address safety, accountability, transparency, and protections against misuse, bias, and harm. Second, I selected AI capacity-building because global participation in AI governance will remain uneven unless countries and institutions have the knowledge, infrastructure, and policy capacity to engage meaningfully. Capacity-building is especially important for developing countries, smaller states, educators, and public institutions that may otherwise be left behind in both governance and innovation. Third, the social, economic, ethical, cultural, linguistic and technical implications of AI are central because AI is not only a technical issue. It affects labor, education, access to information, language representation, cultural inclusion, and social inequality. A strong governance conversation should therefore recognize that AI systems shape human lives differently across contexts and communities. Finally, I chose open-source software, open data and open AI models because openness can support innovation, transparency, collaboration, and broader access, especially when paired with proper safeguards. Open approaches can help reduce dependency on a few dominant actors and make AI development more participatory and globally relevant.

Yes. In addition to the listed themes, I believe several cross-cutting issues deserve more explicit attention. First, environmental sustainability should be treated as a core AI governance issue. AI is often discussed in terms of innovation, safety, and inclusion, but its energy use, water consumption, hardware supply chains, and e-waste implications are increasingly important, especially as large-scale models expand. A global dialogue should consider how AI governance can align with climate and sustainability goals. Second, labour and education transitions need stronger visibility as a distinct governance concern. AI is already reshaping work, skills, teaching, assessment, and professional roles. This is not only an economic issue but also a social and institutional one, affecting workers, students, educators, and public-sector capacity. Governance discussions should therefore address just transition, reskilling, and the future of human expertise. Third, market concentration and infrastructure dependency should be more clearly recognized. Much of the global AI ecosystem depends on a small number of firms controlling compute, cloud infrastructure, chips, foundation models, and data resources. This concentration has implications for competition, sovereignty, affordability, and the ability of smaller or lower-capacity countries to participate meaningfully in AI development and governance. And lastly, governance interoperability and implementation capacity cut across nearly all themes. Many countries are developing AI rules and principles, but a major challenge is how these can work together in practice across jurisdictions, sectors, and levels of capacity. Without attention to interoperability, institutional readiness, and practical implementation, even well-designed principles may remain fragmented or unevenly applied.

In my context, the governance gaps in these thematic areas are affecting Indonesia and Southeast Asia, especially in the education and public-interest sector, in both promising and uneven ways. The main opportunity is that AI is now being taken seriously as a strategic development issue. Indonesia has completed UNESCO's AI Readiness Assessment—the first country in Southeast Asia to do so—and the process identified legal/regulatory, socio-cultural, economic, scientific/educational, and technical/infrastructure dimensions that need coordinated action. Indonesia is also developing a national AI roadmap and has identified priority sectors including education and digital talent. The biggest challenge is that innovation is moving faster than governance capacity. In practice, this affects schools, universities, media, and public institutions that are already using generative AI without clear standards for transparency, safety, data protection, academic integrity, or accountability. ASEAN's expanded guidance on generative AI highlights risks such as inaccurate outputs, disinformation, privacy and confidentiality concerns, deepfakes and impersonation, and the propagation of embedded biases. A second challenge is unequal capacity. Across Southeast Asia, digital government and open-data readiness still lag behind OECD levels: the region averages 0.37 on the OECD Digital Government Index versus 0.61 for OECD countries, and only 23% of high-value datasets are easily accessible to the public versus 47% in OECD countries. This directly affects AI capacity-building, public-sector adoption, and the ability to develop locally relevant and trustworthy systems. At the same time, this creates a major opportunity: if governance is strengthened now, AI could support more inclusive education, better public services, local-language tools, and broader participation in innovation. The most significant need is therefore not only more AI adoption, but stronger institutions, safer guardrails, and wider access so the benefits are distributed fairly.

The AI Dialogue can play a valuable role by serving as a trusted multilateral space where governments, civil society, academia, and technical experts can exchange perspectives on AI governance in a more inclusive and structured way. The UN describes it as an inclusive platform for discussing critical AI issues, sharing best practices, and strengthening international cooperation. Its first contribution can be to build shared understanding. Countries are approaching AI governance differently, but the Dialogue can help identify common concerns such as safety, accountability, transparency, human rights, and capacity gaps. Even without creating binding rules, it can help develop a common language that makes cooperation more practical across regions and institutions. Second, the Dialogue can support policy learning and coordination. Many countries and organizations are already testing frameworks, standards, and regulatory approaches. Bringing these into one forum can help participants compare what is working, avoid duplication, and adapt lessons to different national contexts, especially for countries still building their governance capacity. Third, it can improve inclusion and legitimacy in global AI governance. International discussions are often shaped by a small number of powerful states and technology companies. A meaningful UN-led process can widen participation, particularly for developing countries and underrepresented stakeholders, and help ensure that global governance reflects broader social, developmental, and cultural realities. Finally, the Dialogue can create continuity between discussion and action. If it helps connect annual meetings, expert inputs, capacity-building efforts, and follow-up cooperation, it can become more than a conference. Its strongest role would be to make international AI governance more coordinated, more inclusive, and more responsive to shared global challenges.

The AI Dialogue should build on initiatives that already provide norms, policy guidance, and implementation experience, rather than starting from scratch. First, it should connect with the UNESCO Recommendation on the Ethics of Artificial Intelligence, which is the first global standard on AI ethics adopted by all 194 UNESCO Member States and provides a shared foundation around human rights, transparency, fairness, and human oversight. Second, it should build on the OECD AI Principles and the work of the Global Partnership on AI (GPAI). The OECD Principles are a widely recognized intergovernmental standard for innovative and trustworthy AI, while GPAI helps translate such principles into more practical cooperation among governments, experts, industry, and civil society. Third, the Dialogue should connect with regional mechanisms, including the ASEAN Guide on AI Governance and Ethics and its expanded generative AI guidance. These are valuable because they focus on interoperability, practical governance, and region-specific implementation challenges. The Dialogue's added value would be different from any single one of these processes. As a UN-led global platform, it can bring together countries and stakeholders that are not fully represented in OECD-, GPAI-, or region-specific mechanisms, and create a more inclusive space for developing countries, smaller states, and underrepresented communities. Its strongest contribution would be to connect fragmented efforts: aligning ethical principles, regional guidance, technical lessons, and capacity-building needs into a more coherent international conversation. The Dialogue can also provide continuity through its roadmap, helping transform scattered initiatives into an ongoing process of exchange, coordination, and follow-up.

• Different stakeholders should contribute in complementary ways: governments can share policy experience
• academia and technical experts can provide evidence and risk analysis
• civil society can raise rights, equity, and accountability concerns
• industry can share implementation lessons
• and youth, educators, workers, and affected communities can ground the discussion in lived realities. The Dialogue should combine a high-level governmental segment with multi-stakeholder plenaries, focused thematic sessions, short written inputs, and moderated regional or sectoral consultations. It should also ensure accessible participation, balanced representation, and a clear follow-up mechanism so contributions shape future cooperation, not just one event.

Underrepresented voices include stakeholders from the Global South, smaller and lower-capacity states, Indigenous peoples, minority-language communities, youth, workers, educators, and communities most affected by AI without shaping its rules. UN and UNESCO processes have repeatedly highlighted the need for more meaningful participation by developing countries, young people, and marginalized communities. They could be included through funded participation, multilingual consultations, regional preparatory dialogues, accessible online submission channels, youth and civil society seats in formal sessions, and stronger use of local-language and community-based evidence. Inclusion should be structured, not symbolic, so these perspectives influence agendas, recommendations, and follow-up.

Innovative formats should make the Dialogue more interactive, inclusive, and action-oriented. Useful options include fishbowl discussions, where policymakers, experts, and affected communities rotate into the conversation; scenario workshops on concrete AI governance dilemmas; regional lightning rounds for brief, diverse inputs; and cross-sector problem-solving labs that produce practical recommendations. The UN concept note already points toward plenaries, thematic discussions, and a dedicated multi-stakeholder consultation, which could be strengthened through more participatory design. To deepen engagement, the Dialogue could also use multilingual digital participation tools, structured youth and civil society respondent panels, and short written reaction sessions so contributions shape follow-up outputs, not just live discussion.

One strong example is the UNESCO Recommendation on the Ethics of AI, which gives governments a globally agreed framework grounded in human rights, fairness, transparency, and human oversight. Its added strength is that UNESCO also supports implementation through its Readiness Assessment Methodology (RAM), which helps countries identify institutional and regulatory gaps rather than stopping at high-level principles. A second example is the OECD AI Principles, updated in 2024. They remain useful because they are flexible enough to guide national policy while still promoting common expectations around trustworthy AI, democratic values, and interoperability. This kind of shared baseline is especially important for international cooperation. A third example is the EU AI Act, which is important because it moves from general principles to a concrete legal framework. Its risk-based approach is especially relevant: it focuses stricter obligations on higher-risk AI uses, which is more practical than treating all AI systems the same way. At the regional level, the ASEAN Guide on AI Governance and Ethics and its expanded Generative AI guidance offer a valuable model for contexts where countries have different levels of readiness. These documents are practical, voluntary, and focused on interoperability, while also addressing emerging issues such as accountability, data, and policy responses for generative AI. In my view, the most effective approach is not a single policy, but a layered one: shared principles, readiness assessment, risk-based regulation, and regional guidance adapted to local realities. This combination helps make AI governance both credible and usable, especially for countries and sectors that need concrete tools rather than abstract commitments.