International Organisation for Standardization (ISO)

A successful first Global Dialogue on AI Governance would, above all, create a inclusive platform for open, substantive exchange among all stakeholders. It should enable balanced participation across regions - bringing together perspectives from both developed and developing countries, while actively amplifying voices that are often underrepresented in global governance discussions. Ensuring that socio-technical viewpoints are meaningfully integrated alongside policy, industry, and academic perspectives will be essential to fostering well-rounded and grounded dialogue. Success would also be reflected in the Dialogue's ability to move beyond general statements toward nuanced, structured outcomes. This includes producing clear and thoughtful summaries that capture areas of convergence and divergence, as well as mapping the roles and responsibilities of the diverse actors involved in global AI governance. In particular, highlighting the contributions of International Standards Development Organizations (SDOs), alongside governments, private sector actors, civil society, and the technical community, will be critical. Ultimately, the Dialogue should lay a foundation for continued cooperation by identifying shared priorities and practical pathways forward, including capacity building. Its outcomes should reinforce a collective commitment to advancing AI in ways that are inclusive, open, sustainable, fair, safe, and secure - ensuring that the benefits of AI are broadly distributed and aligned with the public interest.

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

International standards are a trusted mechanism for ensuring the safety, reliability, and trustworthiness of AI systems across countries. Developed through a global, consensus-based process involving more than 175 countries (of which 75% are developing countries), under a one-country-one-vote principle, they provide a uniquely inclusive foundation for operationalizing AI governance worldwide. Standards developed through this approach enables interoperability across jurisdictions, regardless of differing legal or regulatory frameworks. Given their cross-cutting nature, international standards directly support key priority areas identified in General Assembly Resolution 79/325, including safety, trustworthiness, interoperability, accountability and human oversight. They offer practical tools that translate high-level governance principles into implementable solutions applicable across diverse national contexts. In this regard, ISO prioritizes active engagement in advancing and promoting the participation in development of standards and use of international standards as a cornerstone of effective global AI governance. This includes strengthening capacity-building efforts to enhance understanding, adoption, and implementation of standards, particularly in developing countries. By doing so, ISO aims to ensure that all countries can actively participate in and benefit from a harmonized, inclusive, and effective global AI ecosystem.

One important cross-cutting issue that is not sufficiently captured is the environmental impact of AI. The development, training, and deployment of AI systems require significant computational power, which in turn drives high consumption of energy and water resources. Integrating environmental sustainability into AI governance is therefore essential. This includes improving transparency around resource use, encouraging more efficient model design and infrastructure, and aligning AI development with climate goals. Without such considerations, the rapid scaling of AI risks undermining broader sustainability efforts. International standards can play a key role in addressing this challenge. ISO has developed widely adopted standards in areas such as energy management (e.g., ISO 50001), water resource management, and carbon accounting. These frameworks provide practical tools that can be leveraged to measure, manage, and reduce the environmental impacts associated with AI systems. Embedding such standards into AI governance approaches would support more sustainable innovation, ensuring that technological progress is aligned with environmental responsibility.

Standards are widely recognized as essential tools for AI governance. However, significant uncertainty remains around what standards are, what they can realistically achieve, and how they interact with legislation, regulation, and the role of private sector actors. This lack of clarity creates governance gaps, including fragmented approaches, inconsistent implementation, and limited alignment between policy objectives and technical practices. At the same time, global developments are beginning to address these gaps. In response to ongoing international AI governance discussions and the Global Digital Compact, ISO, IEC, and ITU have launched the Seoul Statement. This initiative reflects a coordinated effort by leading standards development organizations to strengthen the role of international standards in AI governance. The Seoul Statement sets out four key commitments: 1. Actively incorporating socio-technical considerations into standards development; 2. Deepening understanding of the relationship between international standards and human rights, recognizing their universal importance; 3. Strengthening an inclusive and dynamic multistakeholder community to support the development and application of AI standards; 4. Enhancing public–private partnerships to build capacity for AI governance.

The AI Dialogue can play a critical role by identifying and promoting trusted platforms where meaningful international cooperation, particularly on socio-technical aspects of AI can take place. This includes mechanisms such as the international standardization system led by ISO and IEC, where international standards and conformity assessment frameworks represent international good practices for ensuring the safety, reliability, and trustworthiness of AI systems, while also facilitating international trade. In addition, the Dialogue can help clarify how different governance tools - standards, regulation, and industry practices - interact, and identify areas where greater alignment is needed. By doing so, it can reduce fragmentation and support more coherent global approaches to AI governance. A key contribution of the AI Dialogue is its ability to foster inclusivity. Ensuring that all voices - across regions, sectors, and levels of development - are heard on equal footing is fundamental to effective international cooperation. The Dialogue can help surface underrepresented perspectives and promote more balanced participation in global governance processes. Finally, the AI Dialogue can highlight the common principles that underpin AI governance efforts worldwide. By revealing shared priorities such as safety, human rights, and trust, it can build mutual understanding and lay the foundation for more coordinated and interoperable approaches to AI governance across jurisdictions.

The AI Dialogue should be designed as a genuinely multistakeholder process. Participation should extend beyond high-level government representatives to include industry, academia, civil society, and international standardization bodies. The inclusion of international standardization organizations is particularly valuable, as they help bridge the gap between high-level policy discussions and the practical, operational tools needed for AI governance. This diversity ensures that the Dialogue reflects the full range of expertise and perspectives required for effective and implementable outcomes. The Dialogue should also clarify and distinguish the roles of UN-affiliated standards bodies and independent international standards development organizations (SDOs), including both their similarities and differences. For example, within the UN system SDOs, members are States, whereas in ISO, members are National Standards Bodies—one per country—typically appointed by governments. Recognizing these institutional differences, as well as their complementarities, is essential for understanding what different SDOs can deliver and for achieving coherent global governance. Structurally, the Dialogue could combine plenary sessions with thematic working groups focused on key issues such as standards, risk management, human rights, and capacity building. Outputs should be practical and action-oriented, including clear recommendations and pathways for collaboration across institutions.

Although international standards developed by ISO/IEC for AI are the most widely used by the market across the globe, ISO and IEC are underrepresented in international governance discussions. It is therefore of great importance that ISO and IEC are represented at all levels, including the highest, during the AI Global Dialogue.

Structured, multistakeholder, and action oriented workshops that focus on practical implementation.

Effective AI governance depends on how well different instruments and actors work together in the AI ecosystem- ranging from international legal frameworks, non-binding principles to national and regional regulation, technical standards, and conformity assessment, and industry practices. When these elements are aligned, they reinforce one another; when they are not, inconsistencies emerge, leading to fragmented oversight, duplication of efforts, and gaps that can erode trust. In this ecosystem, international standards and conformity assessment systems play a distinct operational role. ISO standards, developed by 175 countries one a one-country-one-vote basis, based on consensus, they help implement broad policy goals into measurable and auditable requirements that can be applied throughout the AI lifecycle. However, these tools are only effective when embedded within a broader governance context. Their impact depends on how they are referenced in regulatory frameworks, adopted by industry, and complemented by policy measures. Therefore, a coherent approach to AI governance should therefore recognize the strategic value of standards and conformity assessment as integral components of the overall system. Better coordination across these layers can help reduce fragmentation, strengthen interoperability, and support more consistent and trustworthy AI deployment globally.