ISLT/UCAR

A successful first Global Dialogue on AI Governance would deliver both clear direction and practical momentum. First, it should produce a shared set of guiding principles that reflect global values such as fairness, transparency, accountability, and human rights. While full consensus may be difficult, even a broadly endorsed framework would help align governments, industry, and civil society. Second, success would mean inclusive participation. Voices from the Global South, small states, and underrepresented communities must be meaningfully included—not just present, but influential in shaping outcomes. AI governance must reflect diverse realities, not only the priorities of major powers or tech companies. Third, the Dialogue should lead to concrete next steps, not just discussion. This could include the creation of working groups, timelines for policy development, or pilot collaborations on issues like AI safety, data governance, and ethical standards. Fourth, it should strengthen trust and cooperation between stakeholders. Given the competitive nature of AI development, building channels for dialogue and transparency is essential to reduce fragmentation and risk. Finally, success would involve a commitment to ongoing engagement. AI governance cannot be solved in a single event; the Dialogue should establish itself as a continuous platform for coordination, learning, and adaptation as the technology evolves. In short, the Dialogue would be successful if it moves from ideas to action, ensures global inclusivity, and lays the foundation for sustained international cooperation on AI governance.

• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Interoperability of governance approaches
• Transparency, accountability, and human oversight
• Safe, secure and trustworthy AI

These four priorities are essential to ensure that AI development is both responsible and inclusive. First, safe, secure and trustworthy AI is foundational. Without strong safeguards, AI systems can pose risks to individuals, institutions, and societies. Ensuring reliability, robustness, and security is critical to building public trust and enabling sustainable adoption. Second, AI capacity-building is a key priority, particularly for developing countries and underserved communities. Bridging the knowledge and resource gap is necessary to avoid widening global inequalities. Capacity-building enables more stakeholders to participate meaningfully in AI development and governance. Third, the protection and promotion of human rights must remain central. AI systems can significantly impact privacy, freedom of expression, non-discrimination, and access to opportunities. Embedding human rights principles in AI design and deployment helps prevent harm and ensures that technology serves people. Finally, transparency, accountability, and human oversight are critical for responsible governance. Clear mechanisms are needed to understand how AI systems operate, who is responsible for their outcomes, and how decisions can be challenged or corrected. Human oversight ensures that automated systems do not replace essential human judgment in high-stakes contexts. Together, these priorities support a balanced approach that combines innovation with responsibility, while promoting equitable and ethical AI development globally.

Yes, while the listed themes are comprehensive, several cross-cutting and emerging issues deserve more explicit attention. First, the environmental and climate impact of AI is increasingly urgent. The energy consumption of large-scale AI systems, data centers, and computing infrastructure raises concerns about carbon emissions and resource use. Integrating sustainability into AI development and deployment should be a key priority. Second, data governance and data equity require stronger focus. Questions around who owns, controls, and benefits from data are central to fair AI systems. This includes addressing biases in datasets, ensuring representation of diverse populations, and enabling equitable access to high-quality data, particularly in underrepresented regions. Third, the impact of AI on labor markets and livelihoods is a rapidly evolving issue. Beyond general economic implications, there is a need for proactive strategies to support workforce transitions, reskilling, and protection of vulnerable workers in the face of automation. Fourth, the concentration of AI power among a small number of companies and countries is a critical concern. This raises issues related to competition, digital sovereignty, and equitable participation in AI innovation and governance. Finally, the need for context-sensitive and locally grounded AI solutions is often underestimated. AI systems must be adapted to linguistic, cultural, and socio-economic realities to be effective and inclusive. Addressing these cross-cutting issues would strengthen global AI governance by ensuring it is not only safe and ethical, but also equitable, sustainable, and responsive to diverse global contexts.

Governance gaps in the selected areas—safe and trustworthy AI, capacity-building, human rights, and transparency—are already shaping both challenges and opportunities in our context. One of the most significant challenges is limited AI capacity and infrastructure. Many organizations and entrepreneurs lack access to advanced tools, quality data, and technical expertise. This creates a dependency on external technologies and limits the ability to develop locally relevant solutions. Another key issue is the absence of clear regulatory frameworks. Without well-defined guidelines on transparency, accountability, and human oversight, there is uncertainty around the responsible use of AI. This can slow adoption, reduce trust, and expose users to risks such as bias, misuse of data, or lack of recourse. The protection of human rights is also a concern, particularly in areas such as data privacy, surveillance, and algorithmic bias. Vulnerable populations may be disproportionately affected if safeguards are not embedded early in AI systems. At the same time, there are important opportunities. AI presents strong potential to drive innovation and economic growth, especially by supporting startups and entrepreneurs working on climate, agriculture, health, and digital services. With the right support, local actors can develop solutions tailored to regional needs. There is also an opportunity to leapfrog by adopting responsible AI practices from the outset. By integrating transparency, ethical standards, and human oversight early, the ecosystem can build trust and avoid some of the challenges experienced in more mature markets. Finally, growing awareness around AI creates momentum for capacity-building initiatives and partnerships. Collaboration between governments, academia, and the private sector can help close gaps and build a more inclusive and resilient AI ecosystem. Overall, addressing these governance gaps is essential to ensure that AI development is both beneficial and aligned with local priorities.

The AI Dialogue can play a critical role as a platform for coordination, trust-building, and collective action in advancing international cooperation on AI governance. First, it can help bridge fragmentation by bringing together governments, private-sector actors, academia, and civil society to align on shared principles and priorities. In a rapidly evolving and competitive landscape, the Dialogue can foster convergence around core values such as safety, human rights, and accountability. Second, the Dialogue can promote inclusive participation, ensuring that perspectives from developing countries and underrepresented regions are meaningfully integrated. This is essential to avoid governance models that reflect only the interests of a few actors and to support more equitable global outcomes. Third, it can serve as a space to translate principles into action. By facilitating working groups, partnerships, and pilot initiatives, the Dialogue can support the development of practical tools, standards, and policy approaches that countries can adapt to their contexts. Fourth, the AI Dialogue can strengthen knowledge-sharing and capacity-building by enabling the exchange of best practices, lessons learned, and technical expertise. This can help reduce capability gaps and support more effective implementation of AI governance frameworks. Finally, it can contribute to building trust and transparency among stakeholders. Regular engagement and open dialogue can reduce misunderstandings, encourage cooperation, and create channels for addressing emerging risks collectively. Overall, the AI Dialogue has the potential to move beyond discussion and become a catalyst for coordinated, inclusive, and action-oriented global AI governance.

The AI Dialogue should build on and connect with a range of existing global and regional initiatives to avoid duplication and strengthen coherence in AI governance. Key initiatives include the OECD AI Principles, which provide widely endorsed guidance on trustworthy AI; UNESCO's Recommendation on the Ethics of AI, which offers a global normative framework; and the Global Partnership on AI (GPAI), which advances practical collaboration on responsible AI. Regional efforts such as the European Union AI Act also contribute valuable regulatory approaches. In addition, multi-stakeholder platforms like the Internet Governance Forum (IGF) and technical standard-setting bodies such as the International Organization for Standardization (ISO) play important roles in shaping governance and operational standards. Emerging initiatives like the AI Safety Summit process further highlight growing global attention to AI risks and cooperation. The added value of the AI Dialogue lies in its ability to connect these efforts within a more inclusive and globally representative framework, particularly by amplifying voices from developing countries that are often underrepresented. It can serve as a bridge between norm-setting, technical standards, and implementation, helping translate high-level principles into actionable policies and practices. Moreover, the Dialogue can foster greater alignment and interoperability across initiatives, reducing fragmentation and promoting shared understanding. By creating space for continuous engagement, it can also support coordination, knowledge exchange, and joint action, ensuring that existing efforts reinforce rather than duplicate one another. Ultimately, the AI Dialogue can act as a unifying platform that enhances coherence, inclusivity, and practical impact in global AI governance.

Different stakeholders can contribute to the AI Dialogue by leveraging their unique roles, expertise, and perspectives. Governments can provide policy direction, regulatory frameworks, and national priorities, while also ensuring alignment with international commitments. The private sector can contribute technical expertise, innovation capacity, and practical insights from real-world AI deployment. Academia and research institutions play a key role in evidence-based analysis, risk assessment, and foresight on emerging trends. Civil society organizations can represent public interest, human rights considerations, and the voices of affected communities, ensuring accountability and inclusivity. Finally, international and regional organizations can support coordination, standard-setting, and knowledge-sharing across borders. To be effective, the AI Dialogue should adopt a multi-layered and inclusive structure. This could include: - High-level plenary sessions to set strategic priorities and maintain political momentum -Thematic working groups focused on key issues such as safety, human rights, and capacity-building - Regional consultations to ensure that diverse perspectives, particularly from the Global South, are integrated - Multi-stakeholder roundtables that encourage open exchange between different sectors The format should also emphasize continuity and action. Rather than a one-off event, the Dialogue should function as an ongoing platform with clear timelines, deliverables, and follow-up mechanisms. Digital participation tools can further enhance accessibility and inclusiveness. Finally, the Dialogue should promote transparency and accountability by publishing outcomes, tracking progress, and encouraging stakeholder feedback. Overall, an inclusive, structured, and action-oriented approach will enable meaningful contributions from all stakeholders and ensure that the Dialogue leads to tangible outcomes in global AI governance.

Several important voices remain underrepresented in global discussions on AI governance, which limits the inclusiveness and effectiveness of resulting frameworks. First, Global South countries, particularly from Africa, Latin America, and parts of Asia, are often underrepresented. Their limited participation is due to capacity constraints, resource gaps, and unequal access to decision-making platforms. Yet, these regions are directly affected by AI deployment and bring essential perspectives on development, equity, and local needs. Second, small and medium-sized enterprises (SMEs), startups, and entrepreneurs are frequently overlooked in favor of large technology companies. These actors are key drivers of innovation and can provide practical insights on implementation challenges, market realities, and inclusive solutions. Third, civil society organizations and grassroots communities, including marginalized groups, are often insufficiently included. This includes women, youth, linguistic minorities, and communities most impacted by algorithmic bias or digital exclusion. Their perspectives are critical for ensuring that AI governance reflects real societal impacts. Fourth, non-technical disciplines, such as social sciences, humanities, and local knowledge systems, are underrepresented compared to technical experts. This can lead to governance approaches that overlook social, cultural, and ethical dimensions. To improve inclusion, several actions are needed. These include targeted capacity-building programs, financial and logistical support for participation, and regional consultation mechanisms that feed into global discussions. The use of multilingual platforms and accessible formats can also broaden engagement. Additionally, governance processes should institutionalize multi-stakeholder participation, ensuring that diverse actors are not only consulted but actively involved in decision-making. Strengthening partnerships with local organizations can further help bring grounded perspectives into global debates. Enhancing inclusion will lead to more equitable, legitimate, and effective AI governance outcomes.

To foster meaningful and dynamic engagement, the AI Dialogue should move beyond traditional panels and adopt more interactive, participatory formats. First, multi-stakeholder co-creation labs can bring together policymakers, technologists, civil society, and entrepreneurs to collaboratively design solutions to specific governance challenges (e.g., AI safety frameworks or data governance models). These sessions should be problem-driven and result in concrete outputs. Second, scenario-based simulations and policy labs can help participants explore real-world dilemmas, such as regulating high-risk AI systems or responding to AI-related crises. This approach encourages practical thinking, negotiation, and deeper understanding of trade-offs. Third, "reverse panels" or fishbowl discussions can shift dynamics by allowing traditionally underrepresented voices—such as youth, Global South actors, or community representatives—to lead the conversation, while others listen and respond. This promotes more balanced participation. Fourth, innovation showcases and demo sessions can allow startups, researchers, and practitioners to present real AI solutions, highlighting both opportunities and governance challenges. This grounds discussions in practice rather than theory. Fifth, digital participation platforms (live polls, collaborative documents, and open feedback channels) can enable broader and more inclusive engagement, especially for remote participants. Finally, the Dialogue could include commitment sessions, where stakeholders publicly announce voluntary actions, partnerships, or pilot initiatives, followed by mechanisms to track progress over time. By combining interactive, inclusive, and action-oriented formats, the AI Dialogue can create a more engaging environment that leads to tangible outcomes and sustained collaboration.

Several existing policies and practices offer concrete approaches to advancing effective AI governance. The European Union AI Act is a leading example of a risk-based regulatory framework, classifying AI systems by level of risk and applying proportionate obligations. This approach provides clarity while allowing innovation in lower-risk areas. The UNESCO Recommendation on the Ethics of AI offers a global normative framework grounded in human rights, emphasizing principles such as fairness, transparency, and accountability, and encouraging member states to implement them through national policies. The OECD AI Principles promote trustworthy AI and have been widely adopted, influencing both national strategies and private sector practices. On the operational side, algorithmic impact assessments (AIAs) are increasingly used to evaluate potential risks before deploying AI systems. These assessments help identify bias, privacy concerns, and unintended consequences, and are being integrated into public sector practices in several countries. In the private sector, responsible AI frameworks -including internal ethics guidelines, audit mechanisms, and transparency reports-are becoming more common. These practices support accountability and build public trust. Multi-stakeholder platforms such as the Global Partnership on AI (GPAI) facilitate collaboration on practical challenges, including AI safety and data governance. Additionally, open-source tools and model documentation practices (such as model cards and datasheets for datasets) improve transparency and enable better understanding of AI systems' limitations and intended use. Together, these examples demonstrate that effective AI governance requires a combination of regulation, ethical frameworks, practical tools, and collaborative platforms to address risks while enabling innovation.