Commonwealth Secretariat

A successful first Global Dialogue on AI Governance would contribute to advancing a shared understanding of how AI governance can be effectively implemented across diverse institutional and national contexts. An important outcome would be to strengthen the connection between high-level principles and practical implementation. While significant progress has been made in developing AI governance frameworks, many organisations continue to face challenges in translating these into operational processes, decision-making structures, and system design. The Dialogue could also provide value by facilitating the exchange of practical experiences, including lessons learned and emerging approaches from different sectors and regions. Such exchanges may support the development of adaptable governance models that reflect varying levels of capacity and institutional maturity. Progress towards greater interoperability of governance approaches would also be beneficial. Enhanced alignment, where appropriate, can support cooperation while allowing flexibility for different legal, cultural, and organisational contexts. Finally, establishing a basis for continued engagement beyond the Dialogue, including mechanisms for ongoing collaboration and knowledge-sharing, would help sustain momentum. Overall, success would be reflected in strengthening the link between policy and implementation, contributing to more effective, accountable, and trusted AI governance.

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

These priorities reflect the importance of advancing AI governance from conceptual frameworks towards practical and sustainable implementation. Safe, secure and trustworthy AI remains foundational, as trust is essential for adoption and long-term sustainability. Achieving this requires both technical safeguards and governance structures that support effective oversight and risk management. Interoperability of governance approaches is particularly relevant in a global context. Greater alignment can help reduce fragmentation and support cooperation, while maintaining flexibility to reflect different institutional and regional circumstances. Transparency, accountability, and human oversight are central to ensuring that AI systems operate within defined ethical and risk parameters. In practice, this requires clearly defined roles, responsibilities, and decision-making processes, supported by appropriate monitoring mechanisms. AI capacity-building is also critical. In addition to technical expertise, there is a growing need to strengthen institutional capacity to design, implement, and oversee governance frameworks in a coherent and sustainable manner. Together, these priorities support a more integrated and implementation-oriented approach to AI governance.

A key cross-cutting issue is the operationalisation of AI governance within institutional systems. While principles and frameworks are increasingly well established, challenges remain in embedding these into organisational processes, decision-making structures, and technical environments. This includes integrating AI governance with related domains such as data governance, enterprise risk management, and digital infrastructure, as well as ensuring that governance considerations are incorporated into system design, procurement, and programme delivery. Another emerging issue is the need for lifecycle-based governance. Effective oversight should extend across the full lifecycle of AI systems, from data sourcing and model development to deployment, monitoring, and decommissioning. There is also increasing recognition of the importance of measurable outcomes. Organisations are seeking practical ways to assess whether AI systems operate within acceptable risk thresholds and deliver intended benefits. Addressing these cross-cutting issues could support more effective and sustainable implementation of AI governance across different contexts.

In the context of international organisations, governance gaps in AI present both challenges and opportunities. A key challenge is the gap between high-level principles and operational implementation. While AI-related policies and ethical frameworks are increasingly being developed, practical governance arrangements are still evolving. This can lead to fragmentation, unclear accountability, and inconsistencies in oversight. Interoperability also presents challenges. International organisations often operate across multiple jurisdictions and institutional frameworks, each with different levels of maturity and regulatory approaches. This can make it more complex to develop coherent and aligned governance practices. Capacity constraints are another important consideration. Effective AI governance requires a combination of technical expertise and institutional capabilities, including in data governance, risk management, procurement, and legal review. At the same time, there are significant opportunities. AI has the potential to enhance operational efficiency, improve analytics, and strengthen decision-making across programmes. International organisations are also well placed to facilitate knowledge exchange and contribute to shared governance approaches. Strengthening implementation models, institutional capacity, and interoperability could help ensure that AI is adopted in ways that are safe, accountable, and aligned with organisational mandates.

The AI Dialogue can play an important role as an inclusive multilateral platform for strengthening international cooperation on AI governance. It can support the development of shared understanding by bringing together diverse stakeholders to exchange perspectives in an open and structured manner. This can help identify areas of convergence and promote more coherent approaches while respecting different contexts. The Dialogue also has the potential to support a shift towards implementation. Many stakeholders are seeking practical approaches to operationalising governance, and the Dialogue can facilitate the exchange of experiences, lessons learned, and adaptable models. In addition, it can help ensure that cooperation remains inclusive by reflecting the needs of stakeholders with varying levels of capacity and governance maturity. Finally, the Dialogue can serve as a bridge across existing initiatives, promoting complementarity and identifying opportunities for continued collaboration. Overall, it can contribute to connecting policy discussions with implementation realities in a way that supports effective and inclusive governance.

The AI Dialogue could build upon a range of existing international and multi-stakeholder initiatives that have contributed to the development of AI governance principles and approaches. These include efforts within the United Nations system, UNESCO's Recommendation on the Ethics of Artificial Intelligence, the OECD AI Principles, standards development bodies, and regional initiatives. It would also be valuable to connect with related areas such as data governance, cybersecurity, digital public infrastructure, and enterprise risk management, which are closely linked to implementation. The added value of the AI Dialogue lies in its ability to provide an inclusive multilateral platform that connects these efforts. It can help identify complementarities, promote interoperability, and facilitate the exchange of practical experiences across sectors and regions. In particular, the Dialogue can elevate implementation perspectives by highlighting operational challenges such as accountability, governance integration, and capacity-building. In doing so, it can strengthen the link between normative frameworks and practical application, contributing to a more coherent and actionable approach to AI governance.

Different stakeholders can contribute by bringing complementary perspectives. Governments, international organisations, the private sector, academia, technical communities, and civil society each have distinct roles to play in shaping governance approaches. A balanced structure combining plenary discussions with smaller thematic sessions could support both strategic alignment and practical exchange. Plenary sessions can address broad priorities, while smaller discussions can focus on specific issues. Including case-based sessions and practitioner-focused exchanges may help connect policy discussions with implementation realities. Structured opportunities for cross-sector dialogue could also support mutual understanding. Ensuring continuity beyond the Dialogue, for example through follow-up discussions or knowledge-sharing mechanisms, would help sustain engagement.

Practitioners responsible for implementation within institutions remain underrepresented. This includes those working in ICT, data governance, procurement, risk management, and operational delivery. Voices from developing countries, small states, and institutions with limited capacity are also less visible, as are diverse linguistic and cultural perspectives. These could be included through targeted outreach, hybrid participation, language accessibility, and agenda design that incorporates implementation-focused discussions. Broader inclusion would enhance both the relevance and practicality of discussions.

Interactive formats such as thematic roundtables, small-group discussions, and case-based sessions could support more meaningful engagement. Cross-sector workshops may help bring together different perspectives to address shared challenges. Short practitioner interventions could also provide practical insights from implementation. Hybrid formats should support active participation, including virtual engagement tools and structured contributions. Such approaches may help ensure that discussions are practical, inclusive, and solution-oriented.

Effective AI governance is supported by both normative frameworks and practical implementation approaches. International instruments such as UNESCO's Recommendation and the OECD AI Principles provide important foundations. At the institutional level, effective practices include cross-functional governance structures and integration with existing processes such as procurement, risk management, and data governance. Risk-based approaches and lifecycle oversight can help ensure proportional and consistent governance. Capacity-building and knowledge-sharing also play an important role in strengthening institutional maturity. Overall, approaches that link principles with practical implementation, supported by clear accountability and continuous learning, appear most effective.