The Institute of Electrical and Electronics Engineers, Inc.

A successful Global Dialogue on AI Governance should establish a shared understanding of how to advance inclusive, practical, and globally coordinated approaches to AI governance. As artificial intelligence continues to evolve rapidly and shape economies, societies, and public services, the Dialogue represents an important opportunity to move from high-level principles toward actionable cooperation that enables innovation and supports responsible use. A critical outcome from this perspective would be the identification of areas of common ground across regions and stakeholders, including on transparency, accountability, safety engineering principles and societal safeguards, ethical design considerations, and approaches to deployment. Establishing shared priorities can help reduce fragmentation in governance approaches and support interoperability across national and regional frameworks, while respecting different policy contexts and development needs. Another key measure of success would be strengthening the collaboration with the technical community. Effective AI governance benefits from the participation of governments, industry, standards development organizations, civil society and the technical community. The Dialogue should reinforce mechanisms that bring independent technical expertise and policy perspectives together, from the outset of creating political instruments, helping ensure that governance approaches are both practical and implementable. Recognizing the role of international standards and technical best practices would also contribute to a successful outcome. Standards translate governance objectives into operational guidance, enabling organizations and governments to implement responsible AI in consistent and measurable ways while supporting innovation, enabling cross broader opportunities and promoting global collaboration. Capacity building and inclusive participation are also important outcomes. Ensuring that everyone has opportunities to contribute to and benefit from AI governance discussions will help promote equitable access to AI technologies and their benefits. The first Global Dialogue on AI Governance would be successful if it produces clear next steps — such as ongoing collaboration mechanisms, working groups, and/or shared frameworks — that sustain momentum and strengthen global cooperation. By fostering shared understanding, inclusivity, and actionable collaboration, the Dialogue can help lay the foundation for effective, globally coordinated AI governance.

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

From IEEEs perspective, the following four thematic areas represent priorities for urgent action and active engagement, particularly to ensure that AI governance remains inclusive, practical, and globally interoperable: 1. Safe, Secure and Trustworthy AI Ensuring AI systems are safe, secure, and demonstrably dependable is foundational to building public confidence and enabling sustainable adoption. This includes advancing risk management approaches, robustness, cybersecurity, and reliability throughout the AI lifecycle. Collaborative development of technical standards, best practices, evaluation methodologies, and risk assessments can help operationalize these goals and support consistent implementation. 2. Interoperability of Governance Approaches As countries and regions develop AI governance frameworks, questions arise about how those frameworks relate to one another. Frameworks that complicate cross-border collaboration and data flows often do so because the layer being governed is not clearly specified. At the technical layer, interoperability between systems requires common protocols, shared semantics and specified interfaces. At the governance layer, the question is different: whether frameworks place compatible obligations on the same actors, systems and data flows. These are distinct problems and they call for distinct responses. International standards and conformity assessment mechanisms have a role at the technical layer. At the governance layer, the prior step is to identify where frameworks developed in different forums will apply to the same systems and where their requirements may pull in different directions. That mapping is primarily governance work, not a technical question. Where full alignment is not the right instrument for the context, frameworks can be designed to coexist without producing contradictory obligations. Where alignment is pursued, it should specify which layer it addresses and how compliance can be demonstrated. 3. AI Capacity-Building Participation in AI governance requires more than access to frameworks and standards. It requires the technical capacity to implement, verify and assess compliance with them. For many countries, the gap is not primarily one of knowledge about what governance requires. It is the capacity to test whether systems meet specified requirements, to participate in conformity assessment processes, and to contribute to the standards that shape those requirements. Verification, validation and conformity assessment are not administrative steps; they are the mechanisms through which governance commitments become enforceable in practice. Standards process that are formally open but practically inaccessible, due to cost, language or technical prerequisite, reproduce the asymmetries they are meant to address. Capacity building that enables technical participation in the drafting of standards, not only their implementation, is a governance question, not a secondary concern. In IEEE's experience, the gap is not only one of technical skills or resources. It is also one of presence: in the rooms where standards are drafted, where conformity frameworks are designed, and where the terms of participation are set. 4. Transparency, Accountability, and Human Oversight Openness, transparency and accountability are governance requirements, not aspirational principles. For AI systems, transparency has at least three distinct dimensions: transparency about what a system does, transparency about how it was developed and on what data, and transparency about the institutional arrangements under which it operates. Each dimension requires different instruments and different forms of verification. Conflating them under a single principle produces commitments that cannot be specified or tested. The same applies to accountability: it requires specifying how decisions are made, by whom, and to what end, before it can function as a governance requirement rather than an aspiration.

Several cross-cutting and emerging issues span priorities such as safe, trustworthy AI; interoperability of governance; capacity-building; and transparency and accountability. Addressing these is essential for effective, inclusive, and globally coordinated AI governance. 1. Technical Standards and Operationalization A central challenge is translating high-level principles into practical implementation. Technical standards, best practices, and conformity assessments can operationalize safety, transparency, and accountability while enabling interoperability across jurisdictions. These standards must remain flexible, technology-neutral, and globally relevant. 2. Measurement, Evaluation, and Risk Management Approaches to evaluating AI systems vary widely. Key properties-such as safety, robustness, and performance across populations-lack consistent definitions and standardized methods. Evaluation is foundational, determining whether governance commitments can be verified. Questions around auditing, responsibility, and post-deployment monitoring require technically grounded, globally aligned methodologies. 3. Data Governance and Quality AI system behavior is largely determined by training data. Issues of data quality, provenance, representativeness, and labeling are central to governance, not peripheral. Poor data introduces systemic risks that cannot be fully corrected post-deployment. Privacy and data-sharing frameworks are also critical, as meaningful auditing and cross-jurisdictional evaluation depend on access to appropriate data. 4. Capacity and Global Participation Disparities in resources, expertise, and institutional capacity limit participation in AI governance. Strengthening capacity-building is essential to avoid widening global inequalities. 5. Lifecycle Governance Effective governance must span the full AI lifecycle-from design and data selection to deployment and continuous monitoring. Procurement and system updates are particularly important, as they shape governance outcomes and raise questions about responsibility for evolving systems. 6. Collaboration and Coordination Sustained multi-stakeholder collaboration is necessary to reduce fragmentation and ensure coherent governance approaches. 7. Emerging Technologies Governance must remain adaptable as AI converges with other technologies. Frameworks focused on core governance properties-decision-making, data use, accountability, and verification-are more resilient than those tied to specific technologies. These issues underscore the need for open, coordinated, and technically grounded approaches to AI governance.

Across the priority areas of safe, secure, and trustworthy AI; interoperability; capacity-building; and transparency and accountability, significant governance gaps, challenges, and opportunities remain. Addressing these is essential for effective and inclusive global AI governance. Governance Gaps A key gap lies in operationalizing high-level principles into measurable and implementable requirements. Concepts such as safety, accountability, and oversight are widely endorsed but inconsistently applied across sectors and jurisdictions. At the same time, fragmentation among national and regional frameworks creates uncertainty for organizations operating globally, limiting interoperability and slowing innovation. Capacity disparities further constrain participation. Many countries and organizations lack the technical expertise, infrastructure, and governance tools needed for responsible AI deployment, risking deeper digital divides. In addition, governance often focuses on development and deployment, while post-deployment monitoring—addressing bias, performance, and risk over time—remains underdeveloped. Key Challenges AI technologies are evolving faster than governance frameworks can adapt. Advances such as generative AI and autonomous systems introduce new risks requiring flexible approaches. Measurement and evaluation also remain complex, with limited consensus on methodologies for assessing safety, bias, and reliability. Balancing innovation with risk management presents another challenge. Overly prescriptive frameworks may hinder progress, while insufficient governance can erode trust. Moreover, AI systems operate globally, but governance remains fragmented across jurisdictions, complicating coordination. Emerging Opportunities Momentum is growing around technical standards, risk management frameworks, and best practices that translate governance principles into operational tools. Multi-stakeholder collaboration among governments, industry, and civil society is strengthening alignment. Advances in testing, auditing, and monitoring tools are improving lifecycle oversight, while expanded capacity-building efforts support more inclusive participation. Opportunities exist to develop governance frameworks based on shared principles and system properties rather than specific technologies, enabling adaptability. Strengthening lifecycle governance, conformity assessment, and verification processes can enhance coordination. Leveraging technical standards can bridge policy and implementation, while expanded capacity-building promotes broader participation. Addressing these gaps and challenges presents an opportunity to build a coordinated, inclusive, and effective global AI governance ecosystem that supports both innovation and responsible use.

The Dialogue's most practical contribution is to surface where governance frameworks developed in different forums are producing contradictory obligations on the same actors and systems, and to create the conditions for those contradictions to be addressed before they become entrenched. It can also identify where there is genuine convergence across governance traditions, on the importance of traceability, on the need for independent verification, on the link between data quality and system behavior, and build on that common ground rather than paper over the disagreements that remain. A Dialogue that is technically informed as well as politically representative, and that engages standards bodies and the technical community in substance rather than as implementers of decisions made elsewhere, is better positioned to produce frameworks that hold in practice. Capacity building that enables participation in standards development, not only in governance discussions, is where the development dimension of this work becomes most concrete. The Dialogue can help identify where that participation is formally available but practically inaccessible, and where investment in verification and assessment infrastructure would most strengthen the overall framework. The Dialogue should also: ● serve as a global convening platform that brings together governments, international organizations, industry, standards bodies, and civil society. ● help bridge policy and technical implementation by encouraging collaboration with technical communities and standards development organizations. Finally, the Dialogue can help sustain momentum and coordination by identifying next steps, encouraging ongoing collaboration to advance a coherent, inclusive, and globally coordinated approach to AI governance.

It can build on a growing ecosystem of international initiatives, partnerships, and governance mechanisms by connecting and coordinating these efforts, thus helping to reduce fragmentation, promote interoperability, and accelerate inclusive and practical AI governance. Such as the focus on Ethics of Artificial Intelligence underway such as IEEE P2863TM Recommended Practice for Organizational Governance of Artificial Intelligence provides foundational guidance, while others provide policy guidance such as the OECD AI Principles and those provided by various UN agencies can help operationalize and create alignment across regions. It could serve as a coordination platform to promote coherence among these efforts while identifying shared priorities and practical pathways for implementation. It can play a role connecting standards development and technical governance efforts, as they play a critical role in translating governance principles into technical standards and best practices. The Dialogue can help bridge policy and technical implementation by fostering stronger collaboration between policymakers and technical communities. It can support interoperability and shared learning to assist regional and national governance approaches while addressing different regulatory priorities. Additionally, multi-stakeholder and industry-led initiatives such as the Partnership on AI and open collaboration communities contribute important research, tools, and best practices. Engaging these groups can help ensure that governance approaches remain grounded in practical implementation and evolving technological capabilities. The added value of the Dialogue lies in its ability to connect efforts within a neutral, inclusive, and globally representative forum. It can: ● Promote interoperability across governance approaches ● Identify gaps and duplication across initiatives ● Strengthen capacity-building and knowledge sharing ● Encourage inclusive participation, particularly from developing countries ● Facilitate collaboration between policy and technical communities ● Support ongoing coordination and follow-on actions By building on existing initiatives and strengthening collaboration, the Dialogue can help advance a more coherent, inclusive, and effective global AI governance ecosystem.

AI governance requires meaningful diverse participation through adopting an inclusive, multi-stakeholder structure that enables everyone to contribute their expertise and perspectives in a practical and outcome-oriented manner. Governments contribute policy perspectives, priorities, and experiences, while identifying areas where cooperation and alignment are needed. Participation is critical for advancing interoperable governance and implementation. International organizations help align efforts with global development priorities, share best practices, and facilitate coordination. They support capacity-building, inclusive participation, and knowledge sharing. Industry and private sector stakeholders provide insights on implementation challenges, innovation trends, and operational considerations. Their experience deploying systems helps ensure governance frameworks are practical, scalable, and adaptable to rapidly evolving technologies. Standards development organizations and technical communities play a critical role in translating governance principles into technical standards, best practices, and implementation guidance. Their engagement helps bridge policy objectives with operational and technical realities. The Technical Community contributes evidence-based analysis, emerging technology insights, and independent research. Civil society provides societal impacts, inclusion, and public trust, ensuring that approaches remain human-centered and responsive to diverse needs. To maximize effectiveness, adopt a multi-layered and iterative structure: 1. High-Level Plenary Sessions focused on strategic priorities, shared principles, and identifying areas for cooperation. 2. Thematic Working Groups dedicated and aligned with priority areas such as safety, interoperability, capacity-building, and transparency. 3. Multi-Stakeholder Roundtables for interactive sessions that bring together diverse stakeholders to address implementation challenges and share best practices. 4. Technical and Implementation Sessions with discussions on operationalizing governance approaches, including standards, evaluation frameworks, and capacity-building tools. 5. Regional Consultations for opportunities to incorporate regional perspectives and address diverse governance needs and priorities. 6. Ongoing Collaboration Mechanisms to establish follow-up mechanisms, knowledge-sharing platforms, or pilot initiatives to sustain momentum beyond the Dialogue. Structured, inclusive, and iterative approach will ensure that the Dialogue produces practical outcomes, strengthens international cooperation, and supports inclusive and effective governance.

Several voices and perspectives remain underrepresented in global discussions on AI governance. Ensuring their meaningful inclusion is essential to developing governance approaches that are open, equitable, practical, and globally relevant. To improve inclusion, the Dialogue could: ● Conduct regional and sector-specific consultations ● Establish dedicated stakeholder tracks or advisory groups ● Partner with regional organizations, global standards development organizations and networks for capacity building and training ● Provide multilingual and accessible participation formats ● Create ongoing engagement mechanisms beyond single events By intentionally broadening participation and lowering barriers to engagement, the AI Dialogue can help ensure that global AI governance reflects diverse needs, perspectives, and priorities.

To foster meaningful and dynamic engagement, the Dialogue could adopt innovative, interactive formats that move beyond traditional panel discussions and encourage practical collaboration, inclusive participation, and actionable outcomes. The following engagement formats could help achieve these goals: 1. Multi-Stakeholder Co-Creation Workshops Interactive, facilitated workshops where diverse stakeholders collaboratively develop recommendations, identify governance gaps, or co-design practical solutions. This encourages active participation and allows technical experts, policymakers, industry, and civil society to work together on shared challenges. 2. Scenario-Based Policy Labs Participants work through real-world AI governance scenarios—such as cross-border deployment, risk management, or AI in critical infrastructure—to explore governance implications and identify practical approaches. This translates high-level principles into operational insights to foster deeper understanding. 3. Thematic Roundtables with Rotating Participation Small-group discussions that rotate participants across topics such as safety, interoperability, capacity-building, and transparency. This encourages cross-sector dialogue and allows for engagement with a broader range of perspectives. 4. Lightning Talks and Innovation Spotlights Short, focused presentations highlighting emerging issues, innovative governance approaches, or regional perspectives. This allow a wider range of voices—including underrepresented communities—to share insights and contribute. 5. Technical-to-Policy Bridge Sessions Joint sessions where technical experts and policymakers collaborate to translate governance principles into practical implementation approaches, including standards, evaluation frameworks, and operational guidance. 6. Regional and Community Voices Sessions Sessions designed to elevate perspectives from developing countries, regional organizations, youth, and underrepresented communities to ensure inclusive participation and diverse viewpoints. 7. Interactive Polling and Real-Time Feedback Digital engagement tools that allow participants to provide input during sessions, prioritize issues, and identify areas of consensus to promote dynamic participation and help capture diverse perspectives. 8. Fishbowl Discussions An open dialogue format where participants can rotate into the discussion circle, encouraging spontaneous contributions and more inclusive conversation. 9. Cross-Sector Collaboration Clinics Structured sessions to identify partnership opportunities, capacity-building needs, and collaborative initiatives and generate concrete follow-up actions. 10. Ongoing Dialogue and Follow-Up Mechanisms To sustain engagement, the AI Dialogue could include virtual working groups, knowledge-sharing platforms, and periodic follow-up sessions to continue collaboration and track progress. These innovative engagement formats help create an inclusive, participatory, and outcome-oriented Dialogue that supports meaningful international cooperation and practical governance advancements.

Effective AI governance is emerging through a combination of policies, adopted technical standards, collaborative platforms, and practical implementation approaches. Several examples demonstrate how governance principles can be translated into actionable solutions that address safety, accountability, interoperability, and inclusive participation. Risk-based approaches are increasingly used to tailor governance requirements to the level of potential impact. This approach helps balance innovation and oversight while focusing governance efforts where risks are greatest. Similar risk management approaches are also being developed through standards-based frameworks that support lifecycle governance and continuous monitoring such as IEEE 7001-2021TM Standard for Transparency of Autonomous Systems which establishes measurable, testable levels of transparency, so that autonomous systems can be objectively assessed, and levels of compliance determined. Standards development organizations such as IEEE Standards Association are advancing the development of practical tools to operationalize AI governance. IEEE's efforts have resulted in standards and frameworks for transparency, bias mitigation, risk management, and lifecycle governance. These standards help translate governance principles into measurable and implementable practices. International policy frameworks such as the United Nations Educational, Scientific and Cultural Organization Recommendation on the Ethics of Artificial Intelligence and the Organization for Economic Co-operation and Development AI Principles provide globally recognized guidance for responsible AI development and deployment. These frameworks emphasize human-centered AI, transparency, accountability, and inclusive participation. Organizations are developing practical governance toolkits to support implementation. For example, the National Institute of Standards and Technology AI Risk Management Framework provides guidance for identifying, assessing, and managing AI risks across the lifecycle. These types of tools help organizations operationalize governance principles in real-world settings. Initiatives such as the Global Partnership on Artificial Intelligence and the Partnership on AI provide platforms for collaboration among governments, industry, academia, and civil society. These platforms support knowledge sharing, policy development, and best practice dissemination. Regulatory sandboxes allow organizations to test AI systems under supervised conditions, enabling innovation while addressing governance concerns. These approaches help policymakers better understand emerging technologies and refine governance frameworks. Emerging practices such as AI system documentation, impact assessments, audit mechanisms, and lifecycle monitoring are becoming key governance tools. These practices support accountability and help build trust in AI systems. Capacity-building programs, training initiatives, and knowledge-sharing platforms are expanding to support inclusive participation in AI governance. These efforts help address gaps in technical expertise and governance capabilities. Together, these policies, practices, platforms, and approaches provide concrete examples of how AI governance can move from principles to implementation, supporting safe, trustworthy, and inclusive AI development and deployment.