Future Stars Center for Development and Capacity Building (FSC)

A successful first Global Dialogue on AI Governance should deliver clear, practical, and inclusive outcomes rather than broad statements alone. First, it should produce a shared baseline of principles agreed upon by governments, private sector actors, civil society, UN agencies and international organizations. These principles should address transparency, accountability, safety, human rights, and equitable access, while being flexible enough for different national contexts. Second, the dialogue should result in a roadmap for coordination, outlining how countries and institutions will align existing frameworks, avoid regulatory fragmentation, and promote interoperability. This includes identifying priority areas such as data governance, AI safety standards, and cross-border risks. Third, success would mean establishing multi-stakeholder mechanisms for ongoing engagement. This could include working groups or task forces that continue technical discussions, monitor progress, and ensure that voices from the Global South, including Africa, are meaningfully represented. Fourth, it should generate concrete commitments, not just recommendations. Examples include pledges to adopt ethical AI guidelines, invest in local AI capacity, share knowledge, or support inclusive innovation ecosystems in low- and middle-income countries. Fifth, the dialogue should emphasize capacity building and equity, ensuring that developing countries are not left behind. This includes commitments to funding, training, infrastructure, and access to data and technology. Finally, a successful outcome would include a monitoring and accountability framework, with timelines and indicators to track progress and ensure that commitments translate into action.

• Safe, secure and trustworthy AI
• AI capacity-building
• Transparency, accountability, and human oversight
• Protection and promotion of human rights

1. Safe, Secure, and Trustworthy AI AI contributes by incorporating safety-by-design features such as risk assessment, bias detection, and cybersecurity protections. Continuous monitoring systems can identify harmful behavior or misuse, ensuring AI systems operate reliably and responsibly. Strong safeguards are needed to prevent bias, discrimination, and misuse of AI-especially for vulnerable populations. Clear accountability mechanisms and community-level awareness are key. 2. AI Capacity-Building AI supports capacity-building through accessible learning platforms, automated training tools, and localized content. It can help train communities, especially youth and marginalized groups, in digital and AI skills, enabling them to participate in and benefit from the AI ecosystem. 3. Protection and Promotion of Human Rights AI can help monitor and report human rights violations, improve access to justice, and enhance service delivery in areas like healthcare and education. When designed ethically, it reduces discrimination and supports inclusion, especially for vulnerable populations. 4. Transparency, Accountability, and Human Oversight AI systems can be designed with explainability features that make decisions understandable. Audit trails, documentation, and human-in-the-loop mechanisms ensure that decisions can be reviewed, challenged, and corrected, strengthening accountability and trust.

AI contributes to inclusive humanitarian and development where can be driven locally, while empowering communities to actively shape and benefit from emerging technologies. 5. Equitable Access and Digital Inclusion Ensuring that African and refugee communities have fair access to AI technologies, infrastructure, and digital tools is critical. Without this, AI risks widening existing inequalities rather than reducing them. 6. Data Governance and Protection Communities must have a voice in how their data is collected, used, and shared. Promoting data privacy, ownership, and responsible data practices is crucial, particularly in humanitarian and refugee contexts.

Governance gaps in AI are already shaping both risks and opportunities in Sudan, Uganda and across the African region, particularly within community, humanitarian, and development sectors. One major challenge is the lack of clear regulatory frameworks for AI. This creates uncertainty around data protection, ethical use, and accountability. In contexts involving refugees and vulnerable communities, weak data governance increases the risk of misuse, exploitation, or unintended harm. A second challenge is limited technical capacity and infrastructure. Many institutions and community-based organizations lack the skills, funding, and digital systems needed to safely adopt and manage AI technologies. This widens the gap between global advances and local readiness. Third, there is a growing concern about bias and exclusion. AI systems are often developed using data that does not adequately represent African populations, leading to inaccurate or discriminatory outcomes. Without strong oversight, this can reinforce inequality rather than reduce it. However, these gaps also present important opportunities. AI has strong potential to transform service delivery in sectors like education, healthcare, agriculture, and humanitarian response—through data-driven decision-making, early warning systems, and improved targeting of resources. There is also an opportunity to leverage Africa's young population by investing in AI skills and innovation ecosystems. With the right support, local communities can move from being passive users to active creators of AI solutions. Finally, the current stage of AI development offers a chance for Sudan, Uganda and the region to shape inclusive governance models from the outset—embedding human rights, community participation, and equity into AI systems

The AI Dialogue can play a pivotal role in advancing international cooperation by serving as an inclusive platform for alignment, trust-building, and coordinated action. First, it can help establish a shared global understanding of AI governance principles, bringing together governments, private sector actors, academia, and civil society. This reduces fragmentation and promotes interoperability between national and regional frameworks. Second, the Dialogue can facilitate meaningful inclusion of underrepresented regions, particularly Africa and other parts of the Global South. By amplifying these voices, it ensures that global AI policies reflect diverse realities and do not reinforce existing inequalities. Third, it can act as a space to mobilize partnerships and resources. Countries and organizations can collaborate on joint initiatives such as capacity-building programs, knowledge sharing, and technology transfer, helping bridge the digital and AI divide. Fourth, the Dialogue can support the development of voluntary commitments and cooperative mechanisms, such as joint standards, ethical guidelines, and cross-border risk management approaches. This is especially important for addressing global challenges like misinformation, cybersecurity threats, and misuse of AI. Additionally, it can promote transparency and accountability at the international level by encouraging reporting, peer learning, and the exchange of best practices. Finally, the AI Dialogue can serve as a bridge between policy and practice, ensuring that high-level discussions translate into actionable outcomes, with clear follow-up mechanisms and sustained engagement

The AI Dialogue should build on existing global, regional and local initiatives to avoid duplication and strengthen coherence in AI governance. At the local level: We FSC have designed a project "AI Assistant for Crisis Response" -AACR to support Sudanese Emergency Response Rooms in Darfur—particularly in North Darfur (Tawila) —Online-Offline AI functionality to facilitate the delivery of life-saving humanitarian services and development programs. The project is currently under discussion with OCHA Sudan as well we are looking for donor to copy it in Uganda to support refugees' communities. Regionally, the Dialogue can build on African-led efforts such as the African Union AI Strategy and national digital transformation policies across countries like Uganda. Collaboration with research networks, universities, and innovation hubs is also essential to ground governance in local realities (Patera Data Science currently leading 177 universities for this issue). At the global level, it can connect with frameworks such as the UNESCO Recommendation on the Ethics of AI, the OECD AI Principles, and the work of the Global Partnership on AI. It should also align with digital cooperation efforts under the United Nations, including broader digital governance discussions. Multi-stakeholder partnerships—especially those involving civil society, the private sector, and humanitarian actors—are equally important. These mechanisms already support capacity-building, data governance, and ethical AI use in development contexts. The added value of the AI Dialogue lies in its ability to connect these fragmented efforts into a more coherent global ecosystem. It can act as a bridge between global standards and local implementation, ensuring that principles translate into practice. Additionally, the Dialogue can elevate underrepresented voices, particularly from the Global South, and foster more equitable participation in decision-making. It can also promote coordination, resource mobilization, and knowledge sharing, while encouraging accountability through voluntary commitments and follow-up mechanisms. Ultimately, the AI Dialogue can transform existing initiatives into a more inclusive, aligned, and action-oriented global framework for AI governance.

Different stakeholders can contribute to the AI Dialogue by leveraging their unique roles and expertise, while ensuring inclusive and action-oriented engagement. Governments can provide policy leadership by sharing national strategies, aligning regulatory approaches, and committing to responsible AI frameworks. Private sector and tech companies can contribute technical expertise, innovation, and resources, while committing to ethical standards, transparency, and risk mitigation. Civil society and community-based organizations play a critical role in representing grassroots perspectives, especially from marginalized and underrepresented groups, ensuring that AI governance is inclusive and rights-based. Academia and research institutions can provide evidence-based insights, risk assessments, and independent evaluations of AI systems. International organizations and development partners can facilitate coordination, funding, and capacity-building, particularly for developing countries. Recommendations for Format and Structure The AI Dialogue should adopt a multi-layered and inclusive structure: 1. Plenary Sessions – High-level discussions to set vision, priorities, and political commitment. 2. Thematic Working Groups – Focused groups on key areas such as ethics, data governance, capacity-building, and inclusion, producing concrete recommendations. 3. Regional Consultations – Dedicated spaces for regions like Africa to reflect local priorities and challenges. 4. Multi-stakeholder Roundtables – Interactive sessions encouraging collaboration across sectors. 5. Community Engagement Mechanisms – Inclusion of grassroots voices through consultations, surveys, or hybrid participation. Additionally, the Dialogue should ensure continuity and accountability through: • Clear outputs (action plans, commitments) • Defined timelines and follow-up mechanisms • Open knowledge-sharing platforms Overall, the Dialogue should move beyond discussion to practical collaboration, inclusive participation, and measurable outcomes.

Several voices remain significantly underrepresented in global AI governance discussions, limiting the inclusiveness and effectiveness of outcomes. First, Global South communities, particularly from Africa, are often excluded or underrepresented. Their realities—such as limited infrastructure, informal economies, and humanitarian contexts—are not adequately reflected in global frameworks. Second, refugees, internally displaced persons (IDPs), and crisis-affected populations are rarely included, despite being highly impacted by data collection, surveillance technologies, and digital service delivery systems. Third, grassroots and community-based organizations lack access to global platforms, even though they have deep contextual knowledge of local needs and risks. Fourth, women, youth, and persons with disabilities continue to face barriers to meaningful participation, resulting in AI systems that may unintentionally reinforce inequality and exclusion. Fifth, local technologists, researchers, and innovators from developing countries are often overlooked in favor of actors from more advanced economies. How to Include Them Inclusion requires intentional and structured efforts: Decentralized and regional consultations to capture local perspectives, especially across Africa and other underserved regions. Financial and logistical support (travel grants, stipends, connectivity) to enable participation from marginalized groups. Partnerships with local organizations to channel grassroots voices into global processes. Multilingual and accessible formats to reduce language and technical barriers. Quotas or representation targets to ensure diversity in panels and decision-making spaces. Capacity-building initiatives to empower underrepresented groups to engage effectively in AI policy discussions. By actively integrating these perspectives, AI governance can become more equitable, context-sensitive, and responsive to real-world challenges, particularly in humanitarian and development settings.

To foster meaningful and dynamic engagement, the AI Dialogue should move beyond traditional panel discussions and adopt more interactive, inclusive, and problem-solving formats. First, multi-stakeholder "co-design labs" can bring together governments, technologists, civil society, and affected communities to jointly design AI governance solutions. These labs would focus on real-world case studies, such as humanitarian response, health, or education. Second, regional innovation forums—especially in Africa and other underrepresented regions—can ensure that discussions reflect local realities and generate context-specific recommendations. Third, scenario-based simulation exercises ("AI governance sandboxes") can help participants test policy decisions in simulated environments, exploring risks such as bias, misinformation, or data misuse in real time. Fourth, community listening sessions and digital town halls can enable participation from grassroots actors, including refugees, youth, and marginalized groups, using hybrid and mobile-accessible formats. Fifth, challenge-based hackathons can engage technical communities in developing practical tools for transparency, accountability, and AI safety, directly linking innovation with governance needs. Sixth, peer learning circles and knowledge exchanges can allow countries and institutions to share lessons learned, best practices, and policy experiments in an informal but structured way. Finally, an online global participation platform should complement physical meetings, allowing continuous engagement, idea submission, voting on priorities, and tracking of commitments. These innovative formats ensure that the AI Dialogue is not a one-time consultation but a continuous, participatory, and action-oriented process. They also help bridge the gap between policy, technology, and community realities, making AI governance more inclusive, practical, and responsive to diverse global needs.

Several policies, practices, platforms, and approaches already provide strong foundations for effective AI governance and practical solutions to emerging challenges. At the policy level, the European Union AI Act is a leading example of a risk-based regulatory framework that classifies AI systems by levels of risk and imposes stricter requirements on high-risk applications. It offers a structured approach that balances innovation with safety and accountability. The UNESCO Recommendation on the Ethics of AI provides a global normative framework emphasizing human rights, transparency, fairness, and inclusivity. It is widely adopted and adaptable across different national contexts. The OECD AI Principles also promote responsible AI development, focusing on robustness, accountability, and human-centered values, and are used as a reference point by many governments. At the platform level, initiatives such as the Global Partnership on AI support international collaboration on AI research, capacity-building, and policy experimentation, helping bridge the gap between theory and implementation. Practically, AI audit frameworks and algorithmic impact assessments are increasingly used by governments and organizations to evaluate bias, transparency, and risk before deploying AI systems. Similarly, data protection laws, such as the EU's GDPR, strengthen privacy and user control over personal data. In addition, open-source AI models and shared datasets promote transparency and enable broader participation in AI innovation, especially in developing contexts. Finally, multi-stakeholder governance models, involving governments, private sector, academia, and civil society, ensure more balanced decision-making and help address complex ethical and technical challenges. Together, these examples demonstrate that effective AI governance requires a combination of regulation, ethical standards, technical tools, and inclusive collaboration mechanisms.