Grow Great

A successful first Global Dialogue on AI Governance would move beyond broad principles and produce practical commitments that address inequality, accountability, and inclusion. First, it should establish human rights as the foundation of AI governance, especially for AI used in public services, policing, welfare, education, health, identity, migration, employment, and finance. Public-sector AI should be held to higher standards of transparency, auditability, contestability, and liability. Second, it should create a clear pathway for meaningful Global South participation, not symbolic consultation. African governments, civil society, researchers, youth, marginalized communities, and technical experts should have structured roles in future agenda-setting, standard-setting, and oversight processes. Third, it should recognize economic justice as central to AI governance. This includes fair taxation of AI-driven profits, benefit-sharing from data and language resources, support for community data licenses, and funding mechanisms for local innovation, education, and public-interest AI. Fourth, it should support digital sovereignty and capacity-building by committing resources for African data centers, compute access, cybersecurity, local cloud infrastructure, African language models, and regional research networks. Fifth, the Dialogue should produce a framework for lifecycle governance: responsible design, procurement, deployment, monitoring, evaluation, and redress. AI systems should have trackable outputs, clear accountability, and accessible remedies when harm occurs. Finally, success would mean treating AI as a development and justice issue, not only a technology issue. The outcome should help countries build AI systems that are inclusive, environmentally sustainable, locally relevant, and aligned with human dignity. A strong result would ensure that communities most affected by AI are not just users or data sources, but co-creators and beneficiaries of the AI future.

• Protection and promotion of human rights
• Transparency, accountability, and human oversight
• AI capacity-building
• Social, economic, ethical, cultural, linguistic and technical implications of AI

I selected these priorities because they speak most directly to the urgent governance needs of Africa and the wider Global South. Protection and promotion of human rights is essential because AI systems can affect access to public services, education, health, work, identity, justice, and finance. Governance must prevent discrimination, exclusion, surveillance abuse, and other rights violations. Transparency, accountability, and human oversight are necessary to ensure that AI systems can be explained, audited, challenged, and corrected. People affected by AI decisions should have access to remedies, and responsibility should be clearly assigned to developers, deployers, and public authorities. AI capacity-building is a priority because many African countries still face gaps in compute infrastructure, data governance, research capacity, AI education, and regulatory expertise. Without capacity-building, participation in global AI governance will remain unequal. Social, economic, ethical, cultural, linguistic and technical implications of AI is critical because AI must reflect local realities. This includes African languages, community data rights, economic justice, digital inclusion, environmental impacts, and fair sharing of AI benefits. Together, these priorities support a human-centered and justice-based approach to AI governance that protects rights while enabling inclusive development.

Yes. One cross-cutting issue that needs stronger attention is economic justice and fair distribution of AI value. AI governance should address who owns AI infrastructure, who profits from AI systems, whose data is used, and whether communities receive fair benefits from the data, languages, labour, and knowledge they contribute. Another emerging issue is digital sovereignty. Many African countries depend on foreign cloud infrastructure, proprietary AI systems, and external platforms. Without local compute capacity, data centers, research institutions, and regulatory expertise, countries may struggle to enforce their own laws or build AI systems suited to local needs. A further issue is community data governance. AI systems often depend on cultural, linguistic, and community data, but affected communities rarely have control over how this data is collected, licensed, or monetized. Community data licenses and benefit-sharing models should be explored. The Dialogue should also address public-sector AI standards. AI used by governments should meet higher standards than ordinary commercial AI because it can affect access to rights and essential services. Finally, environmental sustainability should be treated as a core AI governance issue. AI infrastructure consumes energy, water, minerals, and land. Governance should require environmental impact assessment, resource transparency, and sustainable infrastructure planning. These issues cut across human rights, development, accountability, and capacity-building, and are especially important for Africa and the Global South.

Governance gaps in AI are affecting Africa by deepening existing inequalities while also creating opportunities for development, innovation, and rights protection. The most significant challenge is the lack of local AI capacity. Many countries still face limited access to compute infrastructure, data centers, high-quality datasets, AI research funding, and regulatory expertise. This makes African countries dependent on foreign AI systems that may not reflect local laws, languages, cultures, or development priorities. Human rights risks are also increasing. AI systems used in areas such as digital identity, financial services, education, policing, migration, employment, and public service delivery can exclude or discriminate against people if they are not transparent, auditable, and subject to human oversight. Marginalized groups are especially vulnerable where systems are built without local language support, accessibility, or meaningful consultation. There are also economic justice concerns. African data, languages, culture, and digital labour may contribute to global AI development without fair compensation, ownership, or benefit-sharing. Weak governance could allow value to be extracted from communities while profits remain concentrated elsewhere. However, there are major opportunities. With the right governance, AI can improve access to healthcare, education, agriculture, climate adaptation, justice, public services, and local-language information. It can also support African innovation, youth employment, research collaboration, and regional digital markets. The key opportunity is to develop AI governance that is human-centered, rights-based, and development-oriented. This requires investment in AI education, public-interest infrastructure, African language technologies, community data governance, transparent procurement, and regional cooperation. Done well, AI can support digital sovereignty and inclusive development rather than reproduce dependency and exclusion.

The AI Dialogue can play an important role by creating a regular, inclusive space for countries, civil society, technical experts, academia, and affected communities to shape global AI governance together. First, it can help reduce fragmentation by encouraging interoperability between national and regional governance approaches, while still allowing countries to adapt rules to their local contexts. This is especially important for Africa and the Global South, where copy-pasted regulatory models may not address local development priorities, languages, infrastructure gaps, or rights concerns. Second, the Dialogue can promote shared standards on human rights, transparency, accountability, safety, public-sector AI, and human oversight. These standards can help ensure that AI systems are governed responsibly across borders. Third, it can mobilize international cooperation on AI capacity-building. This should include access to compute, data infrastructure, AI education, regulatory expertise, African language technologies, cybersecurity, and public-interest research. Fourth, the Dialogue can support fairer economic arrangements, including benefit-sharing, community data governance, responsible licensing, and global funding mechanisms that help countries participate meaningfully in the AI economy. Finally, it can ensure that those most affected by AI are not excluded from decision-making. The Dialogue should create structured roles for marginalized communities, youth, workers, women, persons with disabilities, and linguistic minorities. Overall, the AI Dialogue can move international cooperation from general principles to practical action: shared standards, funding, technical support, rights protection, and inclusive participation.

The AI Dialogue should build on existing mechanisms rather than duplicate them. Key initiatives include the Global Digital Compact, the UN's Independent International Scientific Panel on AI, UNESCO's Recommendation on the Ethics of AI, the African Union Continental AI Strategy, regional data protection and digital transformation frameworks, and work by OECD/GPAI, standards bodies, civil society networks, academia, and technical communities. UNESCO's Recommendation is especially useful because it already provides a human rights-based ethical framework adopted by UNESCO Member States, while the AU Continental AI Strategy offers an Africa-centered and development-focused approach. The Dialogue should also connect with national AI strategies, data protection authorities, human rights institutions, competition regulators, public procurement bodies, and regional economic communities. These institutions are important because AI governance must be implemented across sectors, not only discussed globally. The added value of the AI Dialogue would be to create a regular, inclusive, UN-based space that connects these initiatives, identifies gaps, and supports practical cooperation. It can help align standards while respecting regional contexts, mobilize funding for AI capacity-building, promote shared approaches to public-sector AI, and ensure that marginalized communities and Global South actors have structured participation. For Africa, the Dialogue can add value by linking global governance discussions to concrete support for compute access, African language technologies, community data governance, digital public infrastructure, regulatory capacity, and public-interest AI. It should become a bridge between global principles and implementation, ensuring that AI governance advances human rights, economic justice, inclusion, and sustainable development.

Different stakeholders should contribute through clearly defined roles. Governments should share national priorities, legal developments, procurement practices, and capacity needs. Civil society should bring human rights, inclusion, community impact, and accountability perspectives. Technical experts and academia should provide evidence on risks, standards, evaluation methods, and emerging capabilities. The private sector should disclose practical deployment challenges, safety practices, environmental impacts, and responsible business models. Marginalized communities, youth, workers, persons with disabilities, women, rural communities, and linguistic minorities should be directly included because they are often most affected by AI systems. The AI Dialogue should be structured as a regular, multi-stakeholder process, not a one-off event. It should include plenary sessions for high-level commitments, technical working groups for specific themes, and regional consultations to ensure local context. Africa and the Global South should have structured representation in agenda-setting, not only participation after decisions are made. The Dialogue should also include public submission processes, community consultations, expert briefings, and youth forums. Outputs should be practical and trackable, including recommendations, implementation roadmaps, capacity-building commitments, funding mechanisms, and follow-up reviews. Working groups could focus on human rights, public-sector AI, capacity-building, economic justice, data governance, language inclusion, environmental sustainability, and AI safety. Each group should include governments, civil society, technical experts, and affected communities. To ensure accountability, the Dialogue should publish summaries of discussions, stakeholder inputs, progress reports, and clear next steps. Its structure should promote cooperation while avoiding domination by a small number of powerful states or companies. The overall format should be inclusive, transparent, regionally balanced, and action-oriented.

Underrepresented voices include African and Global South communities, especially those most affected by AI but least represented in its design and governance. These include rural communities, informal workers, women and girls, youth, persons with disabilities, linguistic minorities, Indigenous communities, migrants, refugees, low-income users, and people without reliable digital access. African language communities are particularly underrepresented. Many AI systems do not adequately support African languages, local knowledge systems, or cultural contexts. This creates exclusion and limits participation in the AI economy. Workers are also underrepresented, including data annotators, content moderators, platform workers, creatives, teachers, health workers, and public servants whose labour and livelihoods are affected by AI. Communities whose data is collected or used to train AI systems are rarely included in decisions about consent, licensing, ownership, or benefit-sharing. These voices can be included through structured participation, not symbolic consultation. The AI Dialogue should create funded participation channels for civil society, youth, community representatives, disability advocates, labour groups, and local-language experts. Consultations should be held regionally and locally, including in African languages and accessible formats. The Dialogue should also support community hearings, youth forums, worker roundtables, and civil society-led submissions. Underrepresented groups should have roles in agenda-setting, working groups, monitoring, and follow-up processes. Inclusion also requires resources. Travel support, translation, accessibility tools, digital access, and institutional funding are necessary so marginalized communities can participate meaningfully. The goal should be to ensure that people most affected by AI are not treated only as users or data sources, but as rights-holders, decision-makers, and co-creators of AI governance.

Innovative engagement formats should make the AI Dialogue participatory, practical, and inclusive rather than limited to formal speeches. One useful format would be regional pre-dialogues and community consultations, especially in Africa and the Global South. These should be held in local languages and accessible formats, with outputs formally feeding into the global agenda. The Dialogue should also include multi-stakeholder working labs where governments, civil society, technical experts, regulators, youth, workers, and affected communities jointly develop practical recommendations on issues such as public-sector AI, human rights, capacity-building, data governance, and economic justice. Another effective format would be AI impact hearings, where communities affected by automated decision-making, data extraction, surveillance, or digital exclusion can share lived experiences directly with policymakers and technical experts. The Dialogue could also use scenario-based exercises to test governance responses to real-world AI risks, such as biased public service systems, AI in elections, harmful surveillance, labour displacement, or exclusion from digital identity systems. A further format would be solution showcases for public-interest AI, African language tools, open models, community data licenses, and inclusive digital infrastructure. This would highlight practical alternatives, not only risks. To encourage accountability, the Dialogue should include commitment sessions where stakeholders announce measurable actions, followed by public progress tracking. Youth assemblies, civil society roundtables, regulator forums, and technical clinics could also help connect high-level policy with implementation. Overall, the most effective engagement formats would be interactive, regionally balanced, accessible, and action-oriented, ensuring that affected communities are not just observers but active contributors to AI governance.

Examples include UNESCO's Recommendation on the Ethics of AI, which provides a human rights-based framework emphasizing dignity, transparency, fairness, and human oversight. This can guide national laws, public-sector AI standards, and institutional ethics processes. The African Union Continental AI Strategy is another important model because it promotes an Africa-centered, development-focused approach to AI governance, linked to ethical, responsible, and equitable AI. It can support regional cooperation, African language technologies, capacity-building, and digital sovereignty. The NIST AI Risk Management Framework offers a practical approach for identifying, assessing, managing, and monitoring AI risks across the AI lifecycle. Its value is that it can be adapted by governments, companies, and public institutions to improve accountability, documentation, testing, and oversight. The Dialogue should also build on data protection laws and authorities, human rights institutions, public procurement rules, and sector regulators. These mechanisms can ensure that AI systems used in health, education, finance, identity, policing, and welfare are lawful, transparent, contestable, and subject to remedies. Other useful approaches include community data licenses, public-interest AI models, regulatory sandboxes with human rights safeguards, algorithmic impact assessments, open data for public-good use, and inclusive AI literacy programmes. Concrete solutions should include higher standards for public-sector AI, mandatory impact assessments for high-risk systems, accessible complaint mechanisms, investment in African data infrastructure, funding for local researchers, and support for African language datasets. Effective AI governance should combine rights protection, technical standards, economic justice, environmental sustainability, and inclusive participation.