Aly

A successful Global Dialogue on AI Governance would advance both practical coordination and deeper conceptual understanding of how AI is reshaping institutional and economic systems globally. First, it would establish a shared analytical framework for understanding AI not only as a technological development, but as an infrastructure shaping economic organisation, decision-making, and power distribution. This would support more coherent and context-sensitive governance approaches across jurisdictions. Second, meaningful progress on interoperability between governance regimes would be essential. From a global perspective, fragmented regulatory approaches risk reinforcing asymmetries between regions. A successful dialogue would explore mechanisms for coordination that recognise diverse institutional contexts while enabling cross-border alignment. Third, capacity-building should be treated as a structural priority rather than a supporting activity. Ensuring that emerging markets can actively participate in shaping and governing AI systems is critical to avoiding the reproduction of existing inequalities within the global digital economy. Finally, the dialogue should create a foundation for sustained engagement between research, policy, and practice. Given the pace of technological change, governance frameworks must remain adaptive, informed by ongoing empirical research and interdisciplinary collaboration.

• Safe, secure and trustworthy AI
• Interoperability of governance approaches
• Transparency, accountability, and human oversight
• Open-source software, open data and open AI models

These priorities reflect a need to understand AI governance as both a technical and institutional challenge. Safe, secure, and trustworthy AI is foundational, particularly as AI systems become embedded within critical economic and financial infrastructures. Trust is not only a technical issue but also an institutional one, shaped by governance, accountability, and regulatory credibility. Interoperability of governance approaches is central to addressing the global nature of AI. From a research perspective, this raises important questions about how different regulatory models interact and how coherence can be achieved without imposing uniformity across diverse economic and political contexts. Transparency, accountability, and human oversight are essential for ensuring that AI systems remain subject to meaningful scrutiny. This includes the ability to understand decision-making processes and to align outcomes with broader societal objectives. Finally, open-source software, open data, and open AI models are critical for enabling broader participation in AI development and governance. They provide opportunities for collaboration, experimentation, and knowledge diffusion, particularly for institutions and regions that may otherwise be excluded from proprietary ecosystems.

A key cross-cutting issue is the relationship between AI and existing economic and monetary systems. AI is increasingly embedded within financial infrastructures, payment systems, and decision-making processes that shape capital allocation and economic activity. This raises important questions about how governance frameworks account for systemic risk, concentration of power, and the evolving role of institutions. Another emerging issue is the spatial dimension of AI development and deployment. There is a risk that AI capabilities become concentrated within a small number of regions and organisations, reinforcing existing global inequalities. Understanding how governance frameworks can address these imbalances requires greater attention to economic geography, infrastructure access, and institutional capacity. Finally, there is a need to consider how governance models can remain adaptive in the face of rapid technological change. Static regulatory approaches may be insufficient, suggesting the importance of iterative, evidence-based frameworks that evolve alongside both technological developments and their societal impacts.

In the financial services sector, governance gaps in AI are already creating both operational challenges and strategic opportunities, particularly in globally interconnected markets such as the United Kingdom and Europe. A key challenge is regulatory fragmentation. Diverging approaches to AI governance across jurisdictions increase complexity for institutions operating across borders, raising compliance costs and creating uncertainty around acceptable use cases. This is particularly relevant as AI systems become more autonomous. The Monetary Authority of Singapore's January 2026 governance framework on Agentic AI represents an important step in addressing oversight, accountability, and control of such systems. However, similar levels of clarity are not yet consistently reflected across other jurisdictions, suggesting a need for broader alignment as agentic AI capabilities evolve. A second challenge relates to the integration of AI governance within existing institutional frameworks. Financial institutions already operate under mature risk, control, and governance structures, and the introduction of AI-specific requirements risks duplication or inconsistency if not carefully aligned. This becomes more complex as agentic systems introduce new forms of decision-making autonomy that may not fit neatly within existing control models. At the same time, there are significant opportunities. Advances in governance frameworks—particularly around transparency, auditability, and model oversight—can strengthen trust in AI systems and enable wider adoption in critical functions. Open-source ecosystems and shared standards also offer the potential to accelerate innovation while improving robustness and collaboration across institutions. From a global perspective, there is also an important opportunity to ensure that emerging markets are active participants in shaping AI governance. Countries such as Nigeria, South Africa, Morocco, and Kenya are increasingly positioning themselves as leaders in digital economies and AI adoption across Africa. However, uneven access to infrastructure, data, and governance capacity risks widening global disparities. Supporting these regions through capacity-building, knowledge exchange, and alignment with emerging governance frameworks will be critical to ensuring a more inclusive and balanced global AI ecosystem.

The AI Dialogue can play a critical role as a convening platform that bridges policy, practice, and research to advance coherent and inclusive approaches to AI governance. First, it can support the development of shared conceptual frameworks that enable different jurisdictions to interpret key governance principles—such as safety, accountability, and transparency—in a consistent yet context-sensitive manner. This is particularly important given the diversity of regulatory approaches emerging globally. Second, the Dialogue can facilitate interoperability between governance regimes by promoting coordination, mutual learning, and the identification of common standards. In a landscape where fragmentation risks slowing innovation and increasing compliance complexity, a structured forum for alignment is essential. Third, it can act as a mechanism for amplifying the voices and priorities of emerging and developing economies. Ensuring that countries across Africa, Asia, and Latin America are active participants in shaping AI governance is critical to avoiding the concentration of technological and regulatory influence in a small number of regions. Finally, the Dialogue can help translate high-level principles into actionable guidance by connecting policymakers with practitioners and researchers. This iterative exchange is necessary to ensure that governance frameworks remain adaptive, evidence-based, and responsive to the evolving capabilities of AI systems, including more autonomous and agentic forms of AI.

The AI Dialogue should build upon and connect existing global and regional initiatives that are already shaping the governance landscape. These include multilateral efforts such as the United Nations' ongoing work on digital cooperation, the OECD AI Principles, and regional regulatory frameworks such as the EU AI Act. In addition, sector-specific initiatives—such as the Monetary Authority of Singapore's January 2026 governance framework on Agentic AI—offer valuable insights into how governance can evolve to address increasingly autonomous systems. The Dialogue should also engage with open-source and technical communities, where standards and practices are often developed in parallel to formal regulation, as well as with emerging digital ecosystems in regions such as Africa, where countries including Nigeria, Kenya, South Africa, and Morocco are actively advancing AI adoption and digital infrastructure. The added value of the AI Dialogue lies in its ability to act as a connective layer across these efforts. Rather than duplicating existing initiatives, it can facilitate coherence, identify gaps, and promote interoperability between frameworks. It can also support knowledge exchange and capacity-building, particularly for regions and institutions that may not have the same resources to engage in multiple parallel initiatives. Ultimately, the Dialogue can help ensure that AI governance evolves as a coordinated, inclusive, and globally relevant system rather than a fragmented set of isolated approaches.

Effective participation in the AI Dialogue requires a structure that reflects the multi-layered nature of AI governance, bringing together policymakers, industry practitioners, researchers, and civil society in a meaningful and coordinated way. A useful approach would be a multi-track format. High-level plenary sessions could focus on strategic alignment and shared principles, while thematic working groups—aligned to priority areas such as interoperability, safety, and open ecosystems—could enable deeper, technical discussions. These working groups should be designed to produce concrete outputs, such as draft recommendations or frameworks. In addition, structured cross-sector dialogues are essential. Bringing together regulators, financial institutions, technology firms, and open-source communities can help bridge the gap between policy intent and practical implementation. This is particularly important in sectors where AI is already embedded within complex governance and risk frameworks. The Dialogue should also incorporate regional perspectives through dedicated sessions that reflect different institutional and economic contexts, ensuring that global discussions are informed by local realities. Finally, continuity is critical. Rather than a one-off event, the Dialogue could establish ongoing communities of practice or task forces that sustain collaboration, track progress, and refine approaches as AI technologies evolve. This would support a more adaptive and iterative model of global AI governance.

Global discussions on AI governance often underrepresent perspectives from emerging and developing economies, as well as practitioners working within complex institutional environments. In particular, there is a need to amplify voices from regions such as Africa and Latin America, where countries are actively advancing digital economies and AI adoption but are not always proportionately represented in shaping global governance frameworks. In Africa, countries including Nigeria, Kenya, South Africa, and Morocco are building momentum in digital innovation and AI. Across Latin America, there is also important progress in digital transformation, public innovation, and AI policy development. Yet these regions are still too often positioned as adopters of governance models designed elsewhere, rather than as active contributors to their design. There is also limited representation from operational practitioners—those responsible for implementing AI within regulated sectors such as finance, healthcare, education, and public administration. Their experience is critical in understanding how governance frameworks work in practice, especially in relation to risk management, compliance, infrastructure constraints, and system integration. To address these gaps, the Dialogue should prioritise inclusive participation through targeted outreach, regional partnerships, multilingual engagement, and practical support for participation from underrepresented groups. This may include hybrid formats, travel support, and collaboration with regional organisations, universities, and policy networks across Africa and Latin America. Ensuring broader representation is not only a matter of equity but also of effectiveness. More inclusive participation will lead to governance frameworks that are more robust, context-aware, and globally relevant.

To foster meaningful and dynamic engagement, the AI Dialogue should adopt interactive, problem-oriented formats that move beyond traditional panel discussions and enable co-creation across stakeholders. One effective approach would be scenario-based workshops, where participants collaboratively examine real-world governance challenges—such as the deployment of agentic AI systems in financial services or public sector decision-making. These sessions would allow stakeholders to test governance frameworks under realistic conditions, identify gaps, and develop practical responses. Another valuable format would be cross-disciplinary policy labs, bringing together regulators, technologists, researchers, and industry practitioners to co-design governance approaches. These labs could focus on translating high-level principles into actionable tools, standards, or guidelines that are adaptable across jurisdictions. To ensure global relevance, the Dialogue could include structured regional exchange sessions, enabling participants from regions such as Europe, Asia, Africa, and Latin America to share perspectives on governance priorities, institutional constraints, and innovation pathways. These exchanges would help surface differences while identifying opportunities for alignment and mutual learning. In addition, hybrid and asynchronous engagement platforms should be integrated into the Dialogue. Pre-event consultations, digital collaboration spaces, and post-event working groups would allow broader participation, particularly from underrepresented regions, and support continuity beyond the event itself. Collectively, these formats would transform the Dialogue into a space for experimentation, collaboration, and sustained engagement—ensuring that AI governance is shaped through inclusive, iterative, and practice-informed processes.

Several emerging policies and practices provide valuable insights into effective AI governance, particularly when examined across jurisdictions and institutional contexts. At the regulatory level, the EU AI Act offers a comprehensive, risk-based framework that categorises AI systems by impact and applies proportionate obligations. This approach has global influence, with extraterritorial implications for organisations operating in the EU market . In Asia, governance models are evolving in distinct but complementary ways. China has developed a state-led approach combining national strategy with targeted regulations on generative AI, algorithmic systems, and content labelling, alongside emerging safety governance frameworks . Japan, by contrast, has adopted a more principles-based and innovation-oriented model, relying on guidance and soft-law mechanisms with limited enforcement, reflecting a different balance between flexibility and control . These contrasting approaches highlight the diversity of governance pathways and the need for interoperability. At a sectoral level, the Monetary Authority of Singapore's January 2026 framework on Agentic AI represents a significant step forward in addressing accountability and oversight for increasingly autonomous systems, offering an early model for future governance developments. Research institutions such as the Alan Turing Institute have also contributed through frameworks on trustworthy AI, ethics, and governance, bridging technical research and policy implementation. In Latin America and the Caribbean, initiatives such as the AI Index (ILIA) developed with ECLAC/CELAC partners demonstrate growing regional engagement. These efforts highlight strong adoption momentum alongside persistent gaps in talent, infrastructure, and governance capacity . Finally, enterprise-level practices-particularly in regulated sectors such as finance-are increasingly embedding AI governance within existing risk, audit, and control frameworks. Alongside this, open-source ecosystems continue to promote transparency, collaboration, and accessibility. Together, these examples demonstrate that effective AI governance is emerging through a combination of regulatory models, institutional practices, and collaborative ecosystems across regions.