NTT DATA Group Corporation

A successful first Global Dialogue on AI Governance should deliver practical and inclusive outcomes that help translate broad principles into usable guidance for governments and stakeholders worldwide. First, the Dialogue should establish a shared foundation for interoperability across AI governance approaches. Rather than seeking a single uniform model, it should help identify common principles, terminology, and core governance elements that can support greater compatibility across jurisdictions. Second, the Dialogue should produce actionable recommendations on safe, secure, and trustworthy AI, including transparency, accountability, and meaningful human oversight. These discussions should be practical enough to support implementation by both public and private sector actors. Third, the Dialogue should prioritize AI capacity-building, especially for developing countries, through cooperation on AI literacy, skills development, institutional capabilities, and access to knowledge and good practices. Capacity-building is essential to ensure that participation in AI governance is truly global and inclusive. Finally, the first Dialogue should create a clear roadmap for continued collaboration beyond the inaugural session, including follow-up workstreams, multistakeholder engagement, and opportunities to share lessons learned from real-world implementation. In our view, success would mean that the Dialogue becomes not only a forum for discussion, but also a trusted global platform that helps bridge governance gaps, reduce fragmentation, and support responsible AI adoption across regions.

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

We selected these priorities because they reflect areas where the Global Dialogue can provide practical value and where international cooperation is most needed. Interoperability of governance approaches is especially important. As AI governance frameworks continue to emerge across regions, greater compatibility among approaches can reduce fragmentation, improve clarity for stakeholders, and support more consistent and responsible adoption of AI globally. We believe the United Nations can play an important role in facilitating practical guidance and common reference points in this area. AI capacity-building is also essential. Inclusive governance requires that countries and stakeholders have the knowledge, skills, and institutional capabilities needed to engage meaningfully. In particular, AI literacy and skills development can help broaden participation and reduce AI divides. We also selected safe, secure and trustworthy AI, together with transparency, accountability, and human oversight, because responsible AI governance must be grounded in implementation. High-level principles are important, but they need to be complemented by practical approaches that help organizations operationalize safety, oversight, and accountability in real-world contexts. In addition to the foregoing, this will specifically support the Protection and promotion of human rights. Together, these priorities can help ensure that AI governance is globally relevant, implementable, and inclusive.

Yes. One important cross-cutting issue is implementation readiness: how to translate governance principles into practical methods, organizational processes, and usable guidance across different sectors and levels of maturity. Many actors do not struggle with the absence of principles, but with how to operationalize them effectively. Another important issue is AI literacy as a foundational enabler of inclusive governance. While capacity-building is listed, broader AI literacy across institutions, workforces, and society deserves stronger visibility because it affects participation, accountability, and the ability to adopt governance measures in practice. A further cross-cutting issue is the need for mechanisms to share implementation experience, including good practices, lessons learned, and practical challenges across jurisdictions and stakeholder groups. A global dialogue can add value not only by discussing norms, but also by enabling structured learning from real-world deployment and governance efforts. Finally, attention should be given to how governance frameworks remain adaptable as AI technologies evolve rapidly. The Dialogue could help promote approaches that are flexible enough to respond to technological change while preserving human rights, accountability, and public trust.

In our sector, one of the most significant challenges is the growing fragmentation of AI governance approaches across jurisdictions. As regulatory and policy frameworks continue to evolve at different speeds and with different definitions, requirements, and expectations, organizations operating across multiple countries and regions face increasing complexity in implementation, compliance, and internal governance alignment. AI should be understood from a global perspective in an increasingly interconnected world, as both its opportunities and challenges apply equally regardless of geography. In this regard, a harmonized governance approach is crucial. A related challenge is the gap between high-level principles and operational practice. Many organizations support the objectives of safe, secure, and trustworthy AI, as well as transparency, accountability, and human oversight. However, practical implementation remains difficult, particularly when organizations need to translate broad governance expectations into internal processes, risk management practices, workforce capabilities, and day-to-day decision-making. Examples and compilations of best practices with real-life cases can represent a very valuable lever to guide all AI stakeholders in the right direction. Capacity-building is another major issue. Differences in AI literacy, institutional readiness, and access to skills and good practices can widen gaps across countries, sectors, and organizations. This affects not only adoption, but also the ability of diverse stakeholders to participate meaningfully in AI governance discussions and implementation. At the same time, these challenges create important opportunities. There is growing momentum for international dialogue, practical guidance, and shared learning. A more interoperable and sharing approach to AI governance could reduce unnecessary fragmentation, improve clarity, and support responsible innovation across regions. Likewise, greater investment in AI literacy and institutional capacity can broaden participation and help ensure that the benefits of AI are more widely shared. From our perspective, the current moment presents an important opportunity to connect governance principles with practical implementation and to strengthen international cooperation in ways that are both inclusive and action-oriented.

The AI Dialogue can play a unique role as an inclusive global platform that helps connect diverse AI governance efforts across regions, sectors, and stakeholder groups. Its value is not to replace existing initiatives, but to help create greater coherence, mutual understanding, and practical cooperation among them. First, the Dialogue can advance international cooperation by fostering interoperability across governance approaches. As AI frameworks continue to develop in different jurisdictions, the Dialogue can help identify shared principles, common terminology, and practical reference points that support greater compatibility while respecting different legal and policy contexts. Second, the Dialogue can strengthen cooperation by providing a space where governments, private sector actors, academia, the technical community, and civil society can exchange implementation experience. Sharing lessons learned, practical challenges, and good practices can help move the conversation from abstract principles toward more actionable governance. Third, the Dialogue can support more inclusive participation in international cooperation by elevating capacity-building, especially for developing countries. AI literacy, skills development, and institutional readiness are essential to ensure that all stakeholders can engage meaningfully in shaping and implementing AI governance. Finally, the Dialogue can help build continuity. By establishing follow-up mechanisms, workstreams, and opportunities for ongoing collaboration, it can support a more sustained and action-oriented model of international cooperation. In this sense, the AI Dialogue can serve as a trusted convening space that helps bridge fragmentation, promote practical alignment, and ensure that international cooperation on AI governance is both inclusive and implementation-focused.

The AI Dialogue should build upon and connect with existing international, regional, and multistakeholder initiatives that are already contributing to AI governance discussions. These include intergovernmental, standards-related, technical, and policy-oriented efforts that have developed useful principles, frameworks, and implementation experience. Rather than duplicating existing work, the Dialogue should serve as a bridge across these efforts so that a coherent organization is established at the international level. Its added value would be to connect initiatives that are often discussed separately, promote mutual learning across regions and sectors, and identify practical points of convergence. In particular, it can help bring together governance discussions, implementation experience, and capacity-building efforts in a more coherent and inclusive manner. The Dialogue could also add value by providing a universal and multistakeholder space within the United Nations system where governments and non-governmental stakeholders can engage on a more equal footing. This is especially important for ensuring that perspectives from developing countries and underrepresented groups are better reflected in global AI governance discussions. Another important contribution would be to elevate practical cooperation. The Dialogue can help connect existing principles and frameworks with real-world implementation challenges, including organizational readiness, AI literacy, and operational governance practices. In this way, the AI Dialogue can complement and improve existing initiatives by strengthening coherence, broadening participation, and creating a trusted platform for structured exchange, practical guidance, and continued collaboration.

Different stakeholders can contribute to the AI Dialogue in complementary ways. Governments can provide policy perspectives and public-interest priorities. The private sector can share practical implementation experience, operational challenges, and lessons from deploying governance measures in real-world contexts. Academia can contribute research, evidence, and independent analysis. The technical community can provide insights on technical feasibility, standards, and system design. Civil society can help ensure that human rights, inclusion, and societal impacts remain central to the discussion. To make these contributions meaningful, the Dialogue should be structured in a way that combines plenary discussions with smaller, focused working sessions. Plenary sessions can help build shared understanding across stakeholder groups, while breakout or thematic sessions can support more detailed discussion on practical issues such as interoperability, capacity-building, transparency, and implementation challenges. The format should also encourage continuity beyond a single event. This could include written submissions, follow-up workstreams, and opportunities to share implementation experience over time. To support inclusive participation, hybrid access, multilingual engagement, and clear documentation of outcomes will be important. In our view, the Dialogue will be most effective if it is designed not only as a forum for statements, but as a structured process for practical exchange, discussion, mutual learning, and continued cooperation among diverse stakeholders. It is important to establish genuine collaboration, with the implementation of real exchanges of information. In this regard, the creation of shared repositories containing documentation to be shared among all stakeholders can be highly beneficial

Several voices remain underrepresented in global discussions on AI governance. These include stakeholders from developing countries, especially those with limited institutional and technical capacity; small and medium-sized enterprises; workers and frontline practitioners affected by AI deployment; speakers of non-dominant languages; and communities whose local, cultural, or social contexts are often insufficiently reflected in global policy discussions. These perspectives should be included not only symbolically, but in ways that enable meaningful participation. This requires reducing practical barriers to engagement, including through hybrid participation options, multilingual access, accessible documentation, and targeted support for stakeholders with more limited resources or institutional capacity. Capacity-building is also essential. AI literacy, policy literacy, and opportunities to engage with governance concepts in accessible ways can help broaden participation. Structured consultations before and after the Dialogue can also help ensure that input is gathered from a wider range of communities, not only from those already active in international policy spaces. In addition, the Dialogue should make room for practical experience from organizations and communities directly affected by AI adoption, including those working in service delivery, education, labor, healthcare, and local public administration. Including these voices can improve the relevance, legitimacy, and real-world usefulness of global AI governance discussions. It could also be worth considering that governments provide effective support and assistance to underrepresented entities so that their voices can be heard in the Dialogue

Meaningful and dynamic engagement would be strengthened by combining formal dialogue with more interactive and practice-oriented formats. One effective approach would be to complement plenary sessions with small, moderated breakout discussions focused on concrete themes such as interoperability, capacity-building, AI literacy, and implementation challenges. Smaller group discussions can encourage more substantive exchange and allow a wider range of participants to contribute. Another useful format would be multistakeholder case-based sessions, where participants share real-world governance experiences, practical challenges, and lessons learned from implementation. This can help ground the discussion in operational reality and encourage structured learning across sectors and regions. Interactive written input processes could also be valuable. For example, collecting short issue papers or practical recommendations before the Dialogue, and then synthesizing them into discussion materials, could help make in-person conversations more focused and inclusive. To broaden participation, hybrid and multilingual engagement formats are essential. Remote participation tools, data sharing repositories, regional preparatory sessions, and accessible summaries can help include stakeholders who may not otherwise be able to engage directly. Finally, the Dialogue could benefit from a follow-up mechanism that keeps participants connected after the event, such as thematic communities of practice, ongoing workstreams, or shared repositories of good practices. Innovative engagement should not only make the event more dynamic, but also strengthen continuity and practical collaboration over time.

At NTT DATA, we believe effective AI governance requires practical approaches that connect high-level principles with implementation at scale. One example we would like to highlight is our global AI literacy and skills development framework for generative AI talent. NTT DATA Group has established a global training framework for approximately 200,000 employees, with four capability levels - Whitebelt, Yellowbelt, Greenbelt, and Blackbelt - to build both foundational literacy and more advanced practical capability. Whitebelt focuses on essential understanding of generative AI, including benefits, risks, and responsible use, while Yellowbelt and above are designed to enable more practical application in delivering value. This practice was shared by NTT DATA with the AI Pact, as part of the Living Repository to foster learning and exchange on AI literacy, and we believe it could be highly beneficial to extend its scope at a global level. We see this kind of structured literacy and skills model as directly relevant to global AI governance discussions. Effective governance depends not only on rules, but also on whether organizations and individuals have the capacity to understand, implement, and operationalize them. AI literacy is therefore a foundational enabler of safe, secure, and trustworthy AI, as well as transparency, accountability, and human oversight in practice. A second important practice is the use of role-based and maturity-based training pathways. By linking learning to different levels of responsibility and practical use, organizations can better align governance expectations with real operational contexts. This helps move AI governance from abstract principles to repeatable internal processes and responsible deployment. We believe the AI Dialogue could add value by encouraging the exchange of such practical models across sectors and regions. In particular, sharing scalable approaches to AI literacy, workforce upskilling, and implementation readiness could make global AI governance more inclusive, interoperable, and actionable.