M-cube Tech Consultants

A successful first Global Dialogue on AI Governance would move beyond discussion and produce clear, practical outcomes that can be acted on across different contexts. First, it should define a shared baseline for responsible AI that is realistic for both high-income and low- and middle-income countries. Right now, most frameworks assume infrastructure and capacity that many regions simply don't have. Second, it should result in concrete commitments, not just principles. For example, commitments to: improve access to affordable data and infrastructure Invest in local AI capacity and education support public-interest AI use cases (healthcare, education, public services) Third, success would mean representation translating into influence. It's not enough for Global South voices to be present — their input should visibly shape the final priorities and outputs. Fourth, the dialogue should produce practical tools or guidance, not just reports. This could include: implementation guidelines for governments design standards that reduce bias and exclusion frameworks that teams (including designers and developers) can actually use Finally, a key outcome would be clear next steps and accountability: what happens after the dialogue who is responsible for what how progress will be tracked Without that, it risks becoming a one-off conversation instead of the start of something that drives real change.

• Safe, secure and trustworthy AI
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Protection and promotion of human rights
• Transparency, accountability, and human oversight

Safe, secure, and trustworthy AI is foundational, but trust cannot exist without transparency, accountability, and meaningful human oversight. In many real-world systems, especially those affecting access to essential services, users are often excluded from understanding or challenging automated decisions. Strengthening transparency and oversight ensures that AI systems remain contestable and aligned with human needs. The social, economic, ethical, cultural, linguistic, and technical implications of AI are particularly urgent in underrepresented contexts. Many current AI systems are not designed with the realities of low-resource environments in mind, which risks reinforcing existing inequalities. Issues such as language exclusion, data affordability, and limited digital access directly affect who benefits from AI and who is left out The protection and promotion of human rights must be embedded across all AI applications, not treated as a separate concern. This includes the right to access information, equitable services, and protection from harm or bias. In practice, this requires designing systems that are inclusive, accessible, and responsive to diverse user needs. Together, these areas reflect a need for governance approaches that are not only high-level, but actionable - ensuring that AI systems are safe, inclusive, and accountable in the contexts where they are deployed.

In South Africa and across much of the Global South, gaps in AI governance are already shaping who benefits from AI and who is excluded. One of the most significant challenges is the mismatch between global AI frameworks and local realities. Many standards for safe and trustworthy AI assume consistent internet access, high data availability, and strong institutional capacity. In practice, issues such as high data costs, limited digital infrastructure, and uneven access to devices mean that AI systems are not equally accessible or usable. There are also growing risks around bias and exclusion, particularly in relation to language and cultural context. Many AI systems are not trained on local languages or datasets, which limits their relevance and can reinforce marginalisation in areas like education, public services, and employment. From a transparency and accountability perspective, there is limited visibility into how automated systems are used in both public and private sectors. This makes it difficult for individuals to understand or challenge decisions that affect them, raising concerns around fairness and human rights. At the same time, there are clear opportunities. AI has strong potential to improve access to essential services, particularly in education, healthcare, and government systems, if designed with local constraints in mind. There is also an opportunity to build more inclusive AI systems from the ground up, rather than retrofitting equity later. Strengthening governance in these areas could enable more equitable innovation, while ensuring that AI systems are both context-aware and accountable to the people they impact.

The AI Dialogue can play a critical role by shifting international cooperation from broad alignment to practical coordination. First, it can help establish a shared baseline for AI governance that is flexible enough to apply across different contexts. Many existing frameworks are shaped by a small number of countries, which creates gaps in relevance and adoption. The Dialogue creates space to co-develop approaches that reflect a wider range of realities, particularly from underrepresented regions. Second, it can act as a bridge between stakeholders who are often disconnected — governments, industry, researchers, and practitioners. Effective AI governance requires these groups to work together, but in practice they operate in silos. The Dialogue can support more continuous collaboration, not just one-off engagement. Third, it can enable knowledge and capacity sharing, especially between regions with different levels of technical and institutional readiness. This includes sharing best practices, tools, and lessons learned, but also supporting local adaptation rather than one-size-fits-all solutions. Fourth, the Dialogue can encourage collective commitments on key issues, such as transparency, accountability, and human rights protections. While not legally binding, these shared commitments can influence national policies, funding priorities, and industry standards. Finally, its value will depend on whether it creates ongoing mechanisms for follow-through. Sustained cooperation requires more than dialogue — it needs continuity, clear priorities, and accountability over time. In this way, the AI Dialogue can move international cooperation from discussion toward more coordinated and context-aware action.