Global AI Governance and Workforce Transformation Policy Observatory

A successful first Global Dialogue on AI Governance should move the global conversation from broad principles toward practical institutional capacity. The Dialogue would be successful if it produces three outcomes. First, it should establish AI governance as an implementation challenge, not only a normative or technical debate. Many countries, schools, employers, and public institutions already understand the importance of safety, inclusion, transparency, and human rights. The harder question is whether they have the capacity to translate these principles into procurement rules, workforce training, risk ownership, data governance, accountability mechanisms, and measurable safeguards. Second, it should create a more inclusive evidence base for AI governance. Global AI governance is still shaped too heavily by governments, frontier technology companies, and a small number of advanced economies. The Dialogue should systematically include evidence from educators, workers, employers, youth leaders, civil society organizations, and institutions in emerging economies. Inclusion should mean more than representation; it should mean giving these actors a structured way to shape priorities, share implementation barriers, and access governance capacity. Third, the Dialogue should identify a small number of shared priorities where international cooperation can produce practical value. These could include institutional readiness indicators, AI literacy for public-interest sectors, responsible adoption standards for education and employment, mechanisms for evaluating AI's labor-market impact, and support for countries and institutions with limited governance capacity. The first Dialogue does not need to solve every AI governance question. Its success should be measured by whether it builds trust, clarifies common priorities, and creates a pathway from high-level consensus to operational governance. The central risk is not only that AI systems may become unsafe; it is that many institutions will adopt AI faster than they can govern it.

Safe, secure and trustworthy AI;AI capacity-building;Social, economic, ethical, cultural, linguistic and technical implications of AI;Interoperability of governance approaches

These four priorities reflect the areas where AI governance most urgently needs to move from principles to institutional practice. Safe, secure and trustworthy AI is foundational because adoption is now moving faster than many institutions' ability to govern it. In education, employment, and public services, trust depends not only on technical safety, but also on clear accountability, human oversight, data protection, procurement standards, and mechanisms for redress. AI capacity-building is equally urgent. Many schools, employers, civil society organizations, and public institutions understand the importance of responsible AI, but lack the operational capacity to implement it. Capacity-building should include AI literacy, workforce training, governance playbooks, risk assessment tools, and support for leaders responsible for deploying AI in real settings. The social, economic, ethical, cultural, linguistic and technical implications of AI are central to my organization's work because AI is already reshaping learning, work, opportunity, and institutional power. These implications cannot be treated separately. For example, language access affects educational inclusion; workforce redesign affects economic mobility; and ethical safeguards affect public trust. Interoperability of governance approaches is essential because AI adoption crosses borders, sectors, and institutional contexts. Fragmented governance can create confusion for educators, employers, technology providers, and policymakers. Interoperability should not mean one universal model imposed everywhere, but a shared basis for mutual learning, compatible standards, and practical cooperation across jurisdictions.

• Yes. One important cross-cutting issue is institutional readiness for AI adoption. Many AI governance discussions focus on principles, risks, technical standards, or national policy frameworks. These are necessary, but they do not fully capture the practical question of whether real institutions are prepared to govern AI once it enters classrooms, workplaces, public agencies, and civil society organizations. Institutional readiness includes leadership accountability, staff capability, procurement practices, data governance, workforce transition planning, human oversight, risk ownership, feedback mechanisms, and evaluation capacity. Without these, even well-designed AI principles may fail in practice. A second emerging issue is the education-to-workforce transition. AI is often discussed separately in education policy, labor policy, and technology governance. In reality, these domains are connected. Education systems shape future AI literacy and opportunity
• employers redesign work and skills demand
• and governments must ensure that transitions are inclusive, safe, and socially sustainable. The Dialogue could help frame AI governance not only as a technology issue, but as a long-term human capability and institutional transition agenda.

In my sector — AI governance, education, and workforce transformation — the most significant governance gap is the distance between rapid AI adoption and institutional readiness. In China and across the broader Asia-Pacific region, schools, companies, and public-interest institutions are experimenting quickly with AI tools. This creates major opportunities: AI can improve access to learning, support personalized education, increase productivity, help workers reskill, and enable organizations to redesign workflows. For emerging economies and resource-constrained communities, AI may also expand access to expertise that was previously unavailable. However, adoption is often faster than governance capacity. Many institutions still lack clear rules on data use, human oversight, procurement, accountability, staff training, and risk evaluation. In education, this can create uncertainty around student privacy, academic integrity, teacher roles, language inclusion, and the quality of AI-supported learning. In the workplace, the challenge is not only job displacement, but unmanaged workflow redesign: employees may use AI without guidance, managers may lack the skills to evaluate productivity gains, and organizations may struggle to distribute benefits fairly. Another challenge is uneven capacity. Large technology firms and advanced institutions can develop internal governance systems, while schools, SMEs, nonprofits, and local public agencies may lack the resources to do so. This risks widening existing gaps between well-resourced and under-resourced institutions. The opportunity is to treat AI governance as a capacity-building agenda. Practical governance tools, readiness indicators, AI literacy program, responsible procurement standards, and sector-specific implementation playbooks could help institutions adopt AI safely and productively. The region also has an opportunity to contribute implementation evidence to the global conversation: not only what AI governance should say in principle, but how it works in real classrooms, workplaces, and organizations.

The AI Dialogue can play a critical role by becoming a bridge between global principles and practical cooperation. First, it can create a trusted multistakeholder forum where governments, International Organisations, companies, educators, employers, civil society, technical experts, youth, and communities from different regions can identify shared governance priorities. This matters because AI governance is currently fragmented across national regulations, voluntary company commitments, technical standards, and sector-specific rules. The Dialogue can help clarify where alignment is possible and where contextual adaptation is necessary. Second, the Dialogue can support interoperability among governance approaches. It should not aim to impose one universal model, but it can help develop common language, shared reference points, and practical guidance on issues such as safety, accountability, data governance, human oversight, capacity-building, and impact assessment. Third, it can elevate implementation evidence from diverse regions and sectors. International cooperation should not be shaped only by advanced economies or major technology companies. The Dialogue can create structured channels for schools, employers, SMEs, local governments, workers, youth organizations, and civil society actors to share what responsible AI adoption looks like in practice. Fourth, it can mobilize capacity-building. Many institutions support responsible AI in principle but lack the tools and capabilities to implement it. The Dialogue can help connect countries and sectors with practical resources, including AI literacy programmes, institutional readiness indicators, governance playbooks, and responsible adoption frameworks. Ultimately, the Dialogue's value will depend on whether it helps the international community move from parallel discussions to coordinated action. Its role should be to build trust, reduce fragmentation, support mutual learning, and help institutions across different contexts adopt AI safely, inclusively, and accountably.

The AI Dialogue should build upon existing international and regional initiatives, while adding a stronger UN-wide bridge between principles, standards, and implementation. Relevant initiatives include the Global Digital Compact, UNESCO's Recommendation on the Ethics of AI and Readiness Assessment Methodology, the OECD AI Principles and OECD.AI Policy Observatory, the GPAI network, the AI Safety Summits and AI Action Summit process and our gaeedu.org These mechanisms already contribute important pieces: principles, technical standards, safety research, policy monitoring, capacity-building, and regional regulation. However, they remain fragmented across institutions, geographies, sectors, and levels of technical capacity. The AI Dialogue can add value in four ways. First, it can serve as a neutral convening space under the UN, connecting governments, International Organisations, companies, educators, employers, civil society, workers, youth, and communities that are often not central in technical or regulatory forums. Second, it can translate existing principles and standards into practical institutional readiness guidance, especially for schools, SMEs, public agencies, and civil society organizations that lack internal AI governance capacity. Third, it can create a structured channel for implementation evidence from diverse regions, including emerging economies. This would help global governance learn not only from advanced regulatory systems and frontier AI companies, but also from institutions facing real adoption constraints. Fourth, it can support interoperability by identifying common reference points across different governance approaches while respecting national and cultural contexts. The Dialogue's added value should not be to duplicate existing initiatives. It should connect them, identify gaps between them, and help ensure that global AI governance becomes more inclusive, practical, and usable for institutions that must govern AI in everyday settings.

Different stakeholders should contribute according to the specific capacities they bring, rather than through general consultation alone. Governments can share policy priorities, regulatory experience, public-sector adoption challenges, and capacity-building needs. International Organisations can help connect existing principles, standards, and development agendas. Technology companies can provide evidence on model capabilities, safety practices, deployment risks, and responsible product design. Employers and CHROs can contribute insight on workforce transformation, job redesign, skills needs, and organizational adoption. Educators and schools can share how AI is affecting teaching, learning, assessment, equity, and child protection. Civil society can surface impacts on rights, inclusion, accountability, and vulnerable communities. Youth, workers, and local communities should be included not only as affected groups, but as evidence providers and agenda shapers. The Dialogue should combine several formats. First, it should include high-level plenaries to identify shared priorities and maintain political visibility. Second, it should create sector-specific working tracks, including education, workforce transformation, public services, health, media, and public administration. These tracks should focus on implementation barriers, not only abstract principles. Third, it should establish regional and local consultation channels so that evidence from emerging economies, schools, SMEs, cities, and civil society organizations can inform the global agenda. The structure should be iterative rather than event-based. The AI Dialogue should function as a continuous learning mechanism that gathers evidence, identifies gaps, connects existing initiatives, and turns global discussion into practical governance support for institutions adopting AI in real settings.

Several voices remain underrepresented in global AI governance discussions. First, educators, school leaders, students, and parents are often treated as affected groups rather than governance stakeholders. Yet education is where AI will shape future literacy, opportunity, assessment, and trust. They should be included through dedicated education tracks, youth consultations, school-based evidence gathering, and practical case studies from different education systems. Second, workers, middle managers, CHROs, SMEs, and frontline employers are not sufficiently represented. AI governance often discusses labor-market impact at a macro level, but less attention is given to how work is actually redesigned inside organizations. These groups should be included through workforce transformation forums, employer-worker roundtables, and sector-specific evidence collection on job redesign, reskilling, productivity, and worker protection. Third, communities from emerging economies and resource-constrained institutions need stronger participation. Their concerns may include infrastructure, language access, affordability, dependency on imported technologies, local capacity, and culturally appropriate governance. Inclusion should involve regional consultations, translation support, travel funding, local research partnerships, and capacity-building before and after consultation. Inclusion should therefore go beyond invitation. It should provide accessible formats, multilingual participation, financial support, practical toolkits, and a clear pathway showing how underrepresented voices influence the Dialogue's priorities and outputs.

The AI Dialogue should move beyond conventional panel discussions and use formats that generate evidence, learning, and practical outputs. One useful format would be AI governance case clinics. Schools, employers, public agencies, SMEs, or civil society organizations could present real AI deployment challenges, such as student data protection, workplace AI adoption, procurement, workforce reskilling, or algorithmic accountability. Multidisciplinary experts could then help diagnose the governance gap and identify practical responses. A second format would be regional implementation labs. These labs could gather stakeholders from different regions to examine how AI governance principles are applied under different infrastructure, language, cultural, economic, and institutional conditions. This would help the Dialogue avoid one-size-fits-all recommendations. A third format would be youth and worker assemblies linked directly to formal agenda-setting. Participants should not only share experiences, but help define priority questions for the Dialogue. Their input should be summarized in official outputs.

Effective AI governance can build on several existing policies and practices. First, UNESCO's Recommendation on the Ethics of AI and its Readiness Assessment Methodology offer a useful model because they connect ethical principles with country-level capacity assessment. This is important because many governance failures come from institutional gaps, not only from lack of principles. Second, the OECD AI Principles and OECD.AI Policy Observatory provide a valuable platform for tracking national AI policies, comparing approaches, and supporting evidence-based policymaking. This type of shared policy infrastructure can help reduce fragmentation. Third, the NIST AI Risk Management Framework offers a practical approach for organizations to identify, measure, manage, and monitor AI risks across the AI lifecycle. Its value is that it can be adapted by public agencies, companies, and other institutions.