ICESCO

A successful first Global Dialogue on AI Governance would not be measured by lofty statements alone, but by tangible alignment, actionable commitments, and credible follow-through. First, shared baseline principles would be essential. Even if countries differ politically, agreeing on core ideas such as safety, transparency, accountability, and human rights would establish a common language for future cooperation. This doesn't require uniform regulation, but it does require mutual recognition of risks. Second, the dialogue should produce practical governance mechanisms, not just declarations. For example, agreement on minimum safety standards for frontier AI systems, incident reporting protocols, or auditing frameworks would signal real progress. Pilot collaborations like cross-border AI safety testing or shared research hubs would be especially valuable. Third, inclusion and legitimacy matter. A successful outcome would involve not just major powers, but also representation from the Global South, industry leaders, researchers, and civil society. AI governance cannot be credible if it is perceived as being dominated by a handful of countries or corporations. Fourth, there should be clear next steps and accountability. This could include establishing a permanent forum, working groups, or timelines for specific deliverables. Without continuity, even strong initial momentum will fade. Finally, success would mean reducing fragmentation. If the dialogue helps prevent incompatible regulatory regimes or escalating "AI nationalism," it will have achieved something significant. Even partial coordination such as interoperability between different governance approaches would be a meaningful win. In short, success looks like moving from conversation to coordination: shared principles, concrete actions, inclusive participation, and a roadmap that ensures the dialogue becomes an ongoing process rather than a one-off event.

• Protection and promotion of human rights
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Safe, secure and trustworthy AI
• Transparency, accountability, and human oversight

Human rights, transparency, and respect for cultural rights are fundamental to responsible AI governance. Human rights ensure that AI remains human-centered. As AI systems increasingly shape decisions in areas like healthcare, employment, and public services, they can amplify discrimination, invade privacy, or restrict freedoms if left unchecked. Anchoring AI governance in universal human rights provides clear ethical boundaries and protects individual dignity. Transparency is equally critical for building trust. AI systems must be explainable and understandable, especially when they affect people's lives. Without transparency, it becomes difficult to identify biases, challenge unfair outcomes, or hold developers accountable. Openness therefore strengthens both accountability and system reliability. Respect for cultural rights ensures that AI reflects global diversity rather than reinforcing dominant perspectives. Since AI is often trained on uneven datasets, it risks marginalizing languages, traditions, and local knowledge. Protecting cultural rights promotes inclusivity and prevents technological homogenization. It is also essential to meaningfully include the Global South. Without their participation, AI governance risks deepening existing inequalities and overlooking local realities. Inclusive dialogue ensures that diverse needs, capacities, and priorities are reflected in global standards. Together, these principles are essential for building AI systems that are trustworthy, equitable, and globally legitimate.

several important cross-cutting and emerging issues often remain underemphasized in AI governance discussions. One key issue is compute and infrastructure inequality. Access to advanced computing power is highly concentrated among a few countries and companies, which risks entrenching global power imbalances and limiting who can meaningfully participate in AI development. Another critical concern is data governance and data justice. Beyond privacy, questions arise about who owns data, who benefits from it, and whether communities especially in the Global South are fairly represented. Without fair data practices, AI systems can reproduce historical inequalities. A particularly important cross-cutting issue is the inclusion of women in and through AI. Women remain underrepresented in AI design and leadership, which directly affects how systems are built. This contributes to algorithmic bias, where AI systems may reflect or amplify gender stereotypes for example in hiring tools, facial recognition, or healthcare applications. Addressing this requires both greater gender diversity in AI development and systematic bias detection and mitigation. Environmental impact is also increasingly important. Training and deploying large AI models consumes significant energy and water resources, raising concerns about sustainability and the uneven global distribution of these costs. Another issue is AI in information ecosystems, including its role in misinformation, deepfakes, and large-scale manipulation, with implications for democracy and public trust. Finally, capacity gaps remain a major challenge. Many countries lack the expertise and resources to govern AI effectively, underscoring the need for international cooperation and knowledge-sharing. Together, these issues show that AI governance must address not only technology, but also inequality, inclusion, and sustainability.

Governance gaps in AI are already shaping both challenges and opportunities across regions and sectors, particularly in Europe and other digitally connected economies. One major challenge is regulatory fragmentation and pace. While frameworks such as the Ryadh Charter or the EU AI Act aim to set clear rules, rapid technological advances often outpace implementation. This creates uncertainty for organizations trying to innovate while remaining compliant, especially smaller actors without legal or technical resources. Another significant issue is inequality in access to AI capabilities. Smaller economies, startups, and public institutions often lack access to advanced compute infrastructure and high-quality datasets. This limits their ability to compete and reinforces dependence on a few dominant global players. Algorithmic bias and inclusion gaps, particularly affecting women and underrepresented groups, also remain a pressing concern. In sectors like hiring, finance, and healthcare, biased systems can reproduce discrimination at scale if not properly audited and governed. At the same time, there are important opportunities. Strong governance frameworks can build trust and drive responsible innovation, positioning regions like Europe as leaders in trustworthy AI. Clear rules can also create a competitive advantage by encouraging the development of high-quality, ethical AI systems. There is also growing momentum for international cooperation and capacity-building, which can help bridge gaps between regions, including better inclusion of the Global South in AI development and governance processes. Finally, AI offers transformative potential across sectors from improving public services to accelerating research and sustainability efforts if governance challenges are addressed effectively. Overall, the key challenge is balancing innovation with oversight, while the main opportunity lies in shaping AI systems that are both competitive and aligned with societal values.

The AI Dialogue can play a pivotal role as a neutral platform for coordination, trust-building, and practical cooperation among diverse stakeholders. First, it can help align global principles by fostering consensus around core values such as human rights, transparency, and accountability, while respecting different regulatory approaches. This reduces fragmentation and supports interoperability between national frameworks. Second, the Dialogue can move beyond discussion to enable concrete collaboration. It can facilitate joint initiatives such as shared safety standards, cross-border testing of AI systems, and coordinated responses to emerging risks like misuse or systemic failures. Third, it can strengthen inclusive participation, ensuring that voices from the Global South, as well as civil society, academia, and industry, are meaningfully involved. This is essential for legitimacy and for developing governance approaches that reflect global realities. The Dialogue can also support capacity-building and knowledge-sharing, helping countries with fewer resources develop the expertise and institutions needed to govern AI effectively. Finally, by establishing ongoing mechanisms—such as working groups, reporting frameworks, or regular convenings, it can ensure continuity and accountability. In sum, the AI Dialogue can transform fragmented efforts into a more coordinated, inclusive, and action-oriented global approach to AI governance.

The AI Dialogue should build upon existing initiatives such as the OECD AI Principles, the Global Partnership on AI (GPAI), the UN Secretary-General's Roadmap for Digital Cooperation, and regional frameworks like the EU AI Act or the Ryadh Charter. It can also connect with research consortia, standards bodies, and civil society coalitions focused on AI ethics, safety, and inclusion. The added value of the AI Dialogue lies in its ability to serve as a truly global, inclusive forum that brings together countries, industry, and civil society on equal footing. Unlike existing mechanisms, it can bridge gaps between regional regulations, facilitate shared safety standards, and address underrepresented voices, particularly from the Global South. By fostering coordination, knowledge-sharing, and concrete pilot collaborations, the Dialogue can accelerate the development of trustworthy, equitable, and widely accepted AI governance practices.