JET Education Services

A successful first Global Dialogue on AI Governance would produce clear, actionable, and globally inclusive outcomes that move beyond high-level principles toward practical commitments. Success would include reaching a shared understanding of the most urgent governance priorities and agreeing on a roadmap for safe, equitable and trustworthy AI that reflects the diverse needs of countries at different stages of digital development. It would demonstrate genuine participation from the Global South, ensuring that their perspectives meaningfully shape frameworks rather than being consulted as an afterthought. A strong outcome would be broad alignment on minimum global expectations for AI safety, transparency, accountability and human oversight, including commitments to risk-assessment processes, auditing mechanisms and safeguarding human rights. Equally important would be tangible progress on capacity-building: agreement on creating accessible funding mechanisms, technical assistance programmes, and opportunities for skills development that enable low- and middle-income countries to implement and benefit from AI governance structures. The Dialogue would also be successful if it produced concrete steps for addressing the social, economic, ethical, cultural and linguistic implications of AI. This could include commitments to support inclusive datasets, locally relevant AI systems, and measures to reduce bias and discrimination. Establishing principles for equitable data access and responsible cross-border data flows would further support global innovation while protecting communities. The success would involve agreeing on a mechanism to sustain momentum beyond the Dialogue, such as follow-up working groups, periodic progress reviews, and a platform for continued cooperation between governments, civil society, academia and industry. If the Dialogue leads to practical commitments, shared standards, strengthened capacities and a clear process for ongoing collaboration, it will lay the foundation for a more coherent, inclusive and future-proof global AI governance ecosystem.

AI capacity-building;Safe, secure and trustworthy AI;Social, economic, ethical, cultural, linguistic and technical implications of AI;Transparency, accountability, and human oversight;

The thematic areas selected I believe closely reflect the priorities that require urgent action and active engagement, especially for the Global South and Africa. Ensuring safe, secure and trustworthy AI is particularly critical in contexts where regulatory capacity, digital infrastructure and oversight mechanisms are still developing. Without strong safeguards, rights-based governance frameworks and risk-mitigation practices, AI deployment in sectors such as education, health, social protection and public services could unintentionally deepen existing inequalities or expose communities to harm. Capacity-building remains one of the most urgent needs across African and Global South regions. Strengthening technical skills, data governance capabilities, institutional readiness and policy expertise is essential for enabling meaningful participation in the global AI ecosystem. Without long-term investment in capacity-building, these regions risk remaining dependent on imported AI systems that are not tailored to local needs, languages or cultural contexts. It is also crucial to address the social, economic, ethical, cultural, linguistic and technical implications of AI in ways that recognize local realities. Africa's linguistic diversity, socio-economic inequalities and cultural richness mean that AI systems that are not designed with contextual sensitivity may reinforce bias, exclude marginalized groups or fail to generate meaningful developmental value. AI must contribute to social wellbeing, equitable economic opportunity and inclusive growth. Finally, transparency, accountability and human oversight are foundational for maintaining public trust. Clear governance mechanisms, explainable systems and meaningful human involvement are particularly important in contexts where automated decision-making can disproportionately affect vulnerable communities.

Several cross-cutting and emerging issues extend beyond the listed themes and warrant dedicated attention in global AI governance discussions. One such issue is the growing concentration of computational power and AI infrastructure in a small number of countries and companies. This "compute divide" has major implications for global equity, sovereignty and innovation, yet is not fully addressed within capacity building or openness discussions. Ensuring fair access to compute, cloud resources and energy efficient AI technologies is becoming a critical governance priority. Another emerging issue is environmental sustainability. The energy demands of large scale AI models, data centres and training cycles contribute to increasing carbon emissions and water usage. Governance frameworks need to consider environmental accountability, green AI standards and incentives for sustainable innovation. Geopolitical dynamics also shape AI development in ways that influence global stability. The use of AI in cyber operations, information manipulation, autonomous systems and military applications raises cross-border risks that are not fully captured within existing themes. Addressing global security implications while preventing the militarisation of AI is essential. AI's impact on labour markets represents another cross-cutting challenge. While economic implications are recognised broadly, the scale and speed of job restructuring, particularly in vulnerable regions, require targeted global strategies for workforce transition, reskilling systems, and social protection mechanisms. Emergent forms of AI, including synthetic media and generative content, introduce new risks related to misinformation, identity manipulation and public trust. Governance responses must include mechanisms for authenticity verification, watermarking, and resilience against large-scale information harms. Another area to consider is the governance of AI supply chains, including hardware, chips, datasets and model components, which raises issues of transparency, traceability and accountability that extend beyond current categories.

Governance gaps and rapid advances in AI are having significant and uneven effects on South Africa and the wider Global South. Limited regulatory capacity and slow development of comprehensive AI governance frameworks mean that new technologies are often adopted faster than they can be governed. As a result, risks related to safety, bias, misinformation and privacy are amplified, particularly in sectors like education, healthcare, policing and social services, where vulnerable populations may be disproportionately affected. A major consequence of these gaps is deepening dependency on foreign developed AI systems. Without local standards, auditing mechanisms or clear accountability structures, South Africa and many Global South countries rely on imported technologies that may not reflect local languages, cultural contexts or socio-economic realities. This increases the likelihood of exclusion, misclassification and discriminatory outcomes. Capacity building remains uneven, and the shortage of specialized AI skills limits the ability of governments and institutions to implement responsible AI practices. This contributes to a widening "governance divide," where regions with fewer resources have less influence over global AI rule making and less ability to adapt or challenge external technologies. The lack of interoperability across governance approaches also affects cross-border cooperation, data flows and regional innovation. For countries working to build digital economies, inconsistent or underdeveloped frameworks hinder their participation in global value chains. Human rights concerns are intensified by the use of AI in surveillance, automated decision-making and public administration where transparency is limited. Without strong oversight, citizens may have little recourse when decisions are opaque or harmful. Other emerging issues, such as technology and infrastructure inequality, environmental pressures from data-centre expansion and rapid labour market disruption, pose additional challenges. These developments risk reinforcing structural inequalities unless addressed through inclusive governance reforms tailored to the needs and realities of South Africa and the Global South.

The AI Dialogue, convened through the United Nations, can play a pivotal role in strengthening international cooperation on AI governance by providing a neutral, inclusive and globally legitimate platform for states, industry, academia and civil society to shape shared approaches. Its greatest value lies in creating space for countries at different stages of technological development, particularly those in the Global South, to participate meaningfully in setting global norms, rather than merely adopting those developed elsewhere. The Dialogue can help build convergence around common principles for safe, trustworthy and human-centred AI while allowing flexibility for national implementation. By clarifying minimum expectations for transparency, accountability, risk management and human oversight, it can reduce fragmentation across governance models and support interoperability between regulatory approaches. A key contribution is facilitating knowledge sharing, capacity building and technical cooperation. The Dialogue can mobilize resources, expertise and partnerships to support countries with limited regulatory capacity, enabling broader participation in setting standard, auditing practices and policy development. It can also promote cooperation on research, multilingual datasets, environmental sustainability, and compute access, areas that no single country can address alone. The Dialogue can advance cooperation on cross-border risks such as misinformation, cyber threats, synthetic media manipulation and the misuse of AI in ways that undermine human rights. Coordinated responses, shared early warning mechanisms and collaborative monitoring can strengthen global resilience. The Dialogue can establish mechanisms for ongoing collaboration, such as follow-up working groups, sharing of case studies, peer review processes, and voluntary reporting frameworks. By building political momentum and institutional continuity, it can help ensure that international AI governance evolves in a coherent, inclusive and future-oriented manner, ultimately promoting equitable and sustainable outcomes for all regions.

The AI Dialogue can build on and connect with several existing global and regional initiatives that are shaping responsible AI development which provide widely recognised frameworks for trustworthy AI and foster technical collaboration among member states. The UNESCO Recommendation on the Ethics of Artificial Intelligence is an important foundational instrument, offering a comprehensive approach and strong emphasis on capacity building, cultural diversity and data governance. At the regional level, the AI Dialogue can draw on the African Union's emerging continental frameworks, including the AU Digital Transformation Strategy and ongoing work toward an African AI strategy. It can also coordinate technical standards and benchmarking processes, as well as initiatives by key international organisations. The added value of the AI Dialogue lies in its ability to integrate these fragmented efforts into a coherent global governance process. Unlike existing initiatives that are often limited by membership, thematic scope or technical focus, the AI Dialogue can offer a universal, inclusive platform where all countries, especially those in the Global South, participate equally in shaping governance norms. It can help bridge the gap between high-level ethical frameworks and practical implementation, aligning standards, risk-management tools and accountability mechanisms across jurisdictions. By convening states, industry, civil society and academia, the Dialogue can address cross-cutting issues such as compute inequality, environmental impacts and cross border risks that no single initiative currently tackles comprehensively. Its convening power, legitimacy and inclusiveness position it to accelerate cooperation, reduce duplication and strengthen global interoperability in AI governance.

Different stakeholders can play complementary roles that strengthen the AI Dialogue's legitimacy, inclusiveness and impact. Governments can contribute by sharing national experiences, identifying regulatory gaps, and committing to interoperable approaches that support global coordination. They can also champion the participation of countries with limited capacity to ensure the Dialogue reflects diverse priorities, not only those of technologically advanced nations. Industry stakeholders can provide technical insights, risk assessments and transparency around model development, deployment practices and supply chains. Their participation is essential for shaping practical governance tools, such as auditing methods, safety evaluations and mechanisms for responsible innovation. However, their involvement should be balanced to avoid undue influence. Civil society organizations can amplify public-interest perspectives, highlight human rights concerns and bring forward the experiences of marginalized communities. Their contributions help ensure that governance discussions remain grounded in social realities and equity considerations. Academia and researchers can support the Dialogue by offering independent evidence, comparative analysis and methodological expertise, especially on emerging risks, evaluation standards and long-term impacts. To maximize these contributions, the AI Dialogue should adopt an inclusive and structured format. A multi-tier approach could be effective: high-level plenary sessions for political alignment; expert working groups for technical and thematic issues; and open consultations that allow broader participation from civil society, youth and community-based organizations. Regional preparatory meetings would ensure context-specific inputs, particularly from the Global South. The Dialogue should also prioritize transparency by publishing agendas, background papers and outcomes, while allowing hybrid participation to reduce barriers. A continuous process could be maintained through periodic reviews, thematic task forces and an online platform for sharing practices, datasets, governance tools and research. Such a structure would ensure that the AI Dialogue remains dynamic, accessible and relevant as the technology evolves.

Several voices and perspectives remain significantly underrepresented in global AI governance discussions. Communities across Africa, often lack the resources, political visibility or technical capacity to participate meaningfully in international forums. As a result, their priorities, such as equitable access to data, compute infrastructure, digital public goods and context-appropriate regulation, are often overshadowed by the interests of wealthier nations and large technology companies. Linguistic and cultural minorities are also underrepresented, especially speakers of low-resource languages whose needs are rarely reflected in global datasets or benchmark standards. This exclusion risks deepening digital inequality and embedding systemic bias within AI systems. Additionally, grassroots civil society, youth groups, educators, persons with disabilities and rural communities are often marginal to formal decision-making processes, even though they are among those most affected by AI's real-world impacts. To address these gaps, global AI governance must adopt more inclusive mechanisms. Dedicated funding for travel, digital access and capacity building would enable participation from low-resource regions. Structured regional consultations, especially in the Global South, can ensure local realities shape global norms rather than being appended at later stages. Multilingual processes, including translation of technical documents and interpretation during dialogues, would support broader engagement. Civil society and community-based organizations should be included through formal seats, not optional side events, and youth participation should be embedded through advisory panels and capacity-building programmes. Creating channels for public-interest technologists, local researchers and practitioners, including those outside major academic centres, would further enrich perspectives. Inclusion requires both representation and influence. Underrepresented voices must not only be present but have the ability to shape agendas, negotiate outcomes and hold institutions accountable. Only then can global AI governance reflect the needs and rights of all communities.

To foster meaningful and dynamic engagement, the AI Dialogue should adopt formats that move beyond traditional plenary discussions and actively involve diverse participants. One effective approach is multi-stakeholder co-creation labs, where governments, industry, civil society and researchers collaboratively develop governance tools, draft recommendations or prototype solutions. These labs encourage multi-stakeholder problem-solving rather than abstract debate from government or academia. Scenario-based simulations can help participants explore AI risks and governance dilemmas in real time, for example, responding to an AI-driven misinformation crisis, a safety incident, or a cross-border data dispute. Such simulations deepen understanding especially for those unfamiliar with AI and reveal practical gaps in existing governance approaches. Rotating regional roundtables, hosted virtually or in-person, would allow communities from the Global South, small states and low-resource regions to shape inputs before global sessions. These roundtables could feed directly into negotiation drafts, ensuring diverse perspectives are integrated early. To broaden access, the Dialogue could incorporate hybrid town-hall forums with open mic segments, enabling grassroots civil society, youth groups, educators and local technologists to share lived experiences. A dedicated Youth Assembly or Future Generations Panel could support intergenerational dialogue and innovation. Thematic rapid-fire briefing sessions which are short, focused presentations by experts could keep participants informed on emerging issues such as environmental impacts, AI inequality or generative AI risks, ensuring discussions remain evidence-based. An interactive digital platform could support continuous engagement between meetings, hosting multilingual resources, deliberation tools, polls, consultation papers and a repository of best practices. Participants could comment on draft recommendations, upload case studies and track progress throughout the process. Another approach that could be adopted is creating participatory drafting rooms, where small groups collectively share, edit and negotiate text in real time which is a dynamic way of building trust and accelerate consensus.

Across South Africa and the wider African region, several policies, platforms and practices demonstrate emerging leadership in responsible AI governance. The African Union has been developing a continental AI strategy focused on ethics, data governance, local capacity-building and equitable innovation. Its Digital Transformation Strategy provides a governance foundation that supports interoperable standards and promotes safeguards for data protection and human rights. At national level, South Africa's work on a national AI policy framework, supported by its earlier Presidential Commission on the Fourth Industrial Revolution, has generated guidelines on ethical innovation, skills development and cross-sector collaboration. The Protection of Personal Information Act (POPIA) also provides a rights-based data governance regime aligned with global standards. Several African countries are piloting regulatory sandboxes that support responsible innovation while allowing governments to test oversight mechanisms in real time. Kenya's emerging AI and data governance initiatives, Rwanda's national AI strategies and Morocco's investment in digital governance frameworks show a growing regional commitment to structured oversight. On the practical side, platforms such as MyMzansi, a digital ID and credentials platform as well as EON Reality in TVET, offer a model for ethical, inclusive, multilingual AI, VR and AR development. Similarly, the AI4D Africa programme has funded community-driven research on AI ethics, agriculture, health and public services, creating locally grounded solutions. In the private and research sectors, South African universities and innovation hubs are advancing AI safety, data governance training and public-interest technology. Civic organizations across the continent also play a growing role in transparency, digital rights advocacy and public engagement. Together these initiatives illustrate how locally driven policies, regional coordination and community-based innovation can strengthen AI governance and offer scalable solutions to global challenges.