Global AI Education and Workforce Transformation Policy Observatory

A successful first Global Dialogue on AI Governance should establish not only a shared vision, but also a practical foundation for sustained international cooperation. To succeed, the Dialogue should recognize that AI governance is not limited to technical standards or frontier risks. It must also address how AI is reshaping education systems, skills development, and workforce transition. These are not secondary issues; they are among the most immediate ways in which AI governance will affect social inclusion, economic opportunity, and public trust. A strong outcome would be the creation of clearer channels between global discussions and local realities. Across different regions, educators, youth leaders, civil society actors, and community-based initiatives are already generating valuable lessons on responsible AI adoption, inclusion, and implementation under uneven conditions. The Dialogue would be most effective if it can bring these grounded experiences into conversation with national and multilateral policy efforts. It would also be valuable for the Dialogue to identify a focused set of actionable priorities, such as inclusive AI adoption in education, support for workforce adaptation, and mechanisms for sharing policy lessons and implementation practices across countries.

Safe, secure and trustworthy AI;AI capacity-building;Protection and promotion of human rights;Transparency, accountability, and human oversight;

AI governance will only be meaningful if it helps societies use AI in ways that are safe, fair, and practical, while also giving people and institutions the ability to adapt to rapid change. We selected safe, secure and trustworthy AI because trust is the basis for any serious adoption. In education, workplaces, and public institutions, people will only use AI with confidence if it is reliable, well-governed, and supported by clear safeguards. Without that, AI risks creating more harm than value. We selected AI capacity-building because governance is not only about setting rules. It is also about whether people and institutions have the knowledge and skills to understand AI, assess its risks, and use it responsibly. This is especially important for educators, workers, young leaders, and public institutions that are already being asked to adapt quickly. We selected protection and promotion of human rights because AI increasingly shapes access to education, employment, information, and opportunity. Governance should ensure that efficiency does not come at the expense of dignity, inclusion, fairness, or people's ability to have a voice in decisions that affect them. We selected transparency, accountability, and human oversight because these are what make governance real in practice. People need to know when AI is being used, who is responsible for the outcomes, and where human judgment remains necessary.

One important cross-cutting issue that deserves more explicit attention is workforce transformation and the challenge of moving from AI pilots to governed adoption. Much of the current discussion focuses either on frontier risks or on isolated use cases, yet in practice one of the most pressing governance questions is how organizations, institutions, and public systems integrate AI into real workflows in ways that are coherent, accountable, and beneficial for people. Many actors can launch pilots, but far fewer can scale them responsibly across teams and functions. This is not only a technical issue, but also a governance issue involving job redesign, skills transition, managerial readiness, workflow ownership, data boundaries, accountability, and human oversight. Without these elements, AI adoption often remains fragmented, with inconsistent use across teams, unclear responsibility, weak safeguards, and limited long-term value. For this reason, workforce transformation should be treated as a distinct and cross-cutting priority, because it sits at the intersection of inclusion, economic opportunity, capacity-building, human rights, and implementation. A successful dialogue on AI governance should therefore pay greater attention to how institutions can move from experimentation to structured, governed, and socially sustainable adoption that improves work while protecting people and strengthening public trust.

In the education and workforce transition sector, especially across China and parts of Asia, the main governance gap is no longer whether AI will be used, but how it is being adopted. The most significant challenge is the growing gap between fast experimentation and slower institutional readiness. Schools, training providers, employers, and public institutions are already testing AI tools, but governance frameworks often lag behind in areas such as data protection, accountability, procurement standards, educator and worker training, and clear rules for human oversight. As a result, adoption can become fragmented: some actors move quickly without sufficient safeguards, while others become overly cautious and avoid adoption altogether. In education, this raises concerns around unequal access, overreliance on tools that are not pedagogically grounded, and insufficient guidance for teachers and students. In the workforce context, the challenge is not only displacement, but also uneven preparedness: many organizations are running pilots, yet few have clear pathways for integrating AI into workflows in ways that improve productivity while protecting workers, clarifying responsibility, and building trust. At the same time, the opportunities are substantial. AI can help expand access to learning, support personalized education, improve public service delivery, and strengthen workforce adaptability if governance is practical and implementation-oriented. There is also a major opportunity for this region to contribute useful lessons to global discussions, because many institutions here are moving quickly and generating real-world experience under conditions of scale, competition, and uneven capacity. The key opportunity is therefore to turn local experimentation into more coherent governance models that are inclusive, credible, and transferable.

The AI Dialogue can play an important role by creating a trusted space where governments, International Organisations, industry, civil society, and local practitioners can exchange not only principles, but also practical lessons from implementation. Its value lies in helping bridge the gap between global frameworks and real-world experience, especially in areas such as education, workforce transition, capacity-building, and responsible adoption. It can also support international cooperation by identifying shared priorities, surfacing common governance challenges across regions, and encouraging the exchange of workable approaches rather than isolated national responses. Most importantly, the Dialogue can help ensure that international AI governance is shaped not only by the most powerful actors, but also by diverse local experiences and public-interest perspectives.

The AI Dialogue should build on existing frameworks that already provide principles, evidence, and coordination channels, including the UN Global Digital Compact and the UN High-Level Advisory Body on AI, UNESCO's Recommendation on the Ethics of AI, and the OECD AI Principles and OECD.AI Observatory. These initiatives have already helped define common norms around trustworthy AI, human rights, transparency, and international cooperation. The added value of the AI Dialogue would be not to duplicate them, but to connect them more effectively: to create a practical bridge between high-level principles and the realities of implementation across countries and sectors. It could add particular value by bringing in more grounded experience from education, workforce transition, public institutions, and local initiatives, and by helping translate fragmented lessons from different regions into shared priorities for cooperation.

Different stakeholders should contribute in different ways. Policymakers can identify priority governance gaps and regulatory needs. Private sector actors can share implementation experience, technical practices, and lessons from moving from pilots to responsible adoption. Individuals and local practitioners can contribute grounded evidence through local projects, policy signal papers, and R&D insights that show how AI is affecting communities in practice. A useful format would combine plenary discussions on major themes with smaller multi-stakeholder working sessions and a clear channel for written inputs. This would allow the Dialogue to connect high-level policy discussion with practical evidence from the field, while making each stakeholder's contribution more distinct and useful.

Regional and community leaders remain underrepresented in many global discussions on AI governance, especially those from lower-income and emerging economies who are leading local projects to address inequality in education, access, and opportunity. These actors often have the clearest understanding of local needs, institutional constraints, and the real social effects of AI, yet their perspectives are often overshadowed by governments, major technology companies, and well-resourced international institutions. Their inclusion would make global dialogue more grounded and more responsive to realities on the ground. They could be better included through dedicated regional representation, support for participation from under-resourced communities, and structured channels for local projects, field evidence, and policy signal papers to feed directly into the Dialogue. This would help ensure that AI governance is shaped not only by those building the technology, but also by those working most closely with the communities affected by it.

A more effective AI Dialogue should go beyond formal speeches and use formats that allow different types of knowledge to surface. One strong option would be short roundtables that mix policymakers, private sector actors, researchers, and local practitioners in the same discussion, so the exchange is not siloed. Another would be case-based sessions, where participants respond to real examples from education, workforce transition, or public service adoption rather than speaking only in abstract terms. The Dialogue could also include regional signal spotlights, where community leaders and local project leads briefly present what they are seeing on the ground, especially from underrepresented regions. Finally, a useful format would be a written input track for short policy signal papers, local project notes, and R&D insights submitted before or after the event, so the Dialogue captures more than what can be said live. Together, these formats would make the process more dynamic, grounded, and useful for future cooperation.

One practical approach is the use of policy signal papers and regional insight platforms that translate local experience into material that larger institutions can actually use. Through our own work in the AI education and workforce transition space, we have been developing a policy observatory model that gathers insights from regional leaders, local projects, and field-based conversations, then consolidates them into short upstreaming papers and analysis designed to inform larger institutional discussions, including those involving the UN, World Economic Forum, and OECD. This kind of approach helps address a common governance gap: global conversations are often strong on principles, but weaker in capturing grounded evidence from diverse local contexts.