National Education Collaboration Trust

A successful first Global Dialogue on AI Governance should produce outcomes that are both principled and implementable. Based on my PhD research on the role of the board in the governance of artificial intelligence, one key measure of success would be whether the Dialogue helps close the gap between high-level ethical commitments and the institutional mechanisms needed to govern AI responsibly. My research found that AI governance often remains underdeveloped because organisations lack clear oversight structures, accountability arrangements, and leadership capacity. First, the Dialogue should establish broad agreement on a core set of governance principles such as fairness, transparency, accountability, safety, privacy, and bias mitigation. These principles featured strongly in my research as essential foundations for responsible AI governance. Second, success would mean advancing practical governance arrangements, not only general declarations. My research argues that effective AI governance requires robust structures, processes, and clearly delegated responsibilities. In that regard, the Dialogue should encourage the development of governance frameworks, oversight bodies, multidisciplinary collaboration, and institutional capacity-building that can translate ethical principles into real-world practice. Third, the Dialogue should be inclusive and globally representative. AI governance cannot be shaped only by powerful states or large technology firms. It must include voices from the Global South, public institutions, civil society, researchers, and communities affected by AI. This is essential if governance is to be legitimate, equitable, and development-oriented. Finally, the Dialogue should create a roadmap for continued cooperation. In my view, its success would lie in whether it helps institutions move from fragmented conversations to shared action: strengthening oversight, building trust, and ensuring that AI serves society in ways that are ethical, accountable, and aligned with the public good.

• Safe, secure and trustworthy AI
• Interoperability of governance approaches
• Transparency, accountability, and human oversight

I would prioritise safe, secure and trustworthy AI; interoperability of governance approaches; and transparency, accountability, and human oversight because these areas speak directly to the governance gaps identified in my PhD research on the role of the board in artificial intelligence governance and ethics. My research found that AI governance often fails not because principles are absent, but because institutions lack clear oversight structures, delegated responsibilities, and practical mechanisms to translate principles into action. Safe, secure and trustworthy AI is urgent because institutions need confidence that AI systems are reliable, robust, ethical, and aligned with societal values. My research highlights safety, security, fairness, privacy, and bias mitigation as core minimum principles of an effective AI governance framework. Transparency, accountability, and human oversight are equally important because AI should never operate without clear responsibility and review. My research argues that governance frameworks must define who is accountable, how authority is delegated, and how boards and management exercise oversight over AI systems. Interoperability of governance approaches matters because AI governance cannot remain fragmented across institutions, sectors, and jurisdictions. My work emphasises the need to align AI governance with wider corporate governance, IT governance, risk governance, and legal frameworks, so that governance systems are coherent rather than siloed.

The AI Dialogue can play an important role in advancing international cooperation by creating a credible global platform for shared norms, practical learning, and coordinated action. From my perspective, its value lies not only in setting principles, but in helping institutions translate those principles into workable governance arrangements. In this context, the AI Dialogue can support international cooperation in four ways. First, it can help build a shared understanding of minimum governance principles such as fairness, transparency, accountability, safety, security, and human oversight. Second, it can encourage interoperability between national and institutional governance approaches, reducing fragmentation across jurisdictions. Third, it can promote capacity-building, especially for developing countries and under-resourced institutions, so that AI governance is not limited to those with advanced technical and regulatory capacity. Finally, it can create space for inclusive, multi-stakeholder engagement involving governments, academia, civil society, and the private sector.

The AI Dialogue should build on, and connect with, existing initiatives rather than duplicate them. In my view, key foundations include the UNESCO Recommendation on the Ethics of AI, which already provides a globally endorsed normative baseline on human rights, transparency, accountability and human oversight; the OECD AI Principles and OECD.AI/GPAI ecosystem, which offer practical policy guidance, measurement tools and expert collaboration; the UN Secretary-General's High-level Advisory Body on AI, which has already outlined proposals for global institutional cooperation; and the Council of Europe Framework Convention on AI, which is the first international legally binding treaty in this field. The added value of the AI Dialogue would be to serve as a bridging platform: linking normative frameworks, policy guidance, technical standards, and national governance experiences into a more coherent global process. It can provide political legitimacy under the UN, create space for stronger participation from the Global South, and help reduce fragmentation across existing initiatives.

• Different stakeholders can contribute to the AI Dialogue by bringing different but complementary strengths: governments can provide policy leadership and create enabling regulatory environments
• the private sector can share technical knowledge, innovation trends, and practical lessons from implementation
• academia can offer evidence-based research and independent analysis
• civil society can ensure that human rights, inclusion, and public interest remain central
• and affected communities can bring lived experience that helps ground discussions in real social impacts.

Perspectives from the Global South are still not adequately reflected, despite the fact that many countries will be asked to govern, procure, and live with AI systems developed elsewhere. Global debates often privilege regulatory, commercial, and technical priorities from the most resourced regions, while giving less attention to development realities, institutional capacity constraints, and local public-interest concerns. The UN has emphasized the need for an inclusive platform that brings all countries and stakeholders into the conversation.

The Dialogue should therefore be structured as an inclusive, multi-stakeholder platform that combines high-level plenary sessions with smaller thematic working groups and practical roundtables. This would allow both broad strategic discussion and focused engagement on specific governance issues. It should also include opportunities for written submissions, regional consultations, and follow-up mechanisms so that participation is not limited to those physically present. Most importantly, the format should support continuity, clear outputs, and action-oriented cooperation rather than a once-off exchange of views.