Online Side Event Summary: Global Dialogue on AI Governance (Geneva, 2026)

In connection with the first UN Global Dialogue on AI Governance, we hosted an official online side event on July 6, 2026, focused on current AI standardization practices in the African region, as well as the challenges and opportunities for improved coordination and cooperation, both regionally and internationally. 

Key areas discussed include the necessity for AI standards and their roles in governing frontier AI risks, continental standardization efforts, realities and implementation gaps, and ways to improve regional coordination and international cooperation, among other important discussion points.

https://www.youtube.com/watch?v=I9tnRwsW19M

Key takeaways from the discussion are:  

• Standards are Necessary for Robust Oversight of AI: The discussion emphasized the need for robust AI standards to ensure AI capabilities remain safe, as traditional software frameworks are inadequate to address the unique risks posed by frontier models, including cyber and bio threats, as well as the potential for loss of human control.

  
  

• Continental AI Standardization Effort: Highlighted continental AI standardization efforts include the recent creation of the IT Technical Committee 046 at the African Organization for Standardization (ARSO) level. Discussed candidate priorities for the committee include developing standards for AI development processes (including developer codes of practice and standards for data and tools) and creating standards for use cases to promote the safe application of AI technology. 

  
  

• Realities & Implementation Gaps: The discussion highlighted that, while AI standards are essential for measuring compliance and enabling government regulation, a significant gap exists: most African nations are still drafting basic AI strategies and policies rather than enforceable regulations. Furthermore, imposing heavy, unharmonized global frameworks (like the EU AI Act) creates severe compliance bottlenecks for early-stage African startups that often consist of just a few developers. To bridge these implementation gaps, the proposed approach includes designing standards that reflect local realities, distinct rules for upstream providers versus downstream developers, and building significant technical capacity within regulatory bodies to ensure that standards can actually be understood and enforced.

• Obstacles to Producing Harmonized Standards and Adoption for Africa: The discussion emphasized that the biggest obstacle to producing harmonized AI standards that member states will actually adopt is the exclusion of African data, research, and local context during the initial "problem framing" stage of development. While African nations frequently participate in international technical committees and hold voting rights, they are often relegated to voting on solutions for problems defined by developed nations rather than co-designing standards for their own realities. To ensure true regional ownership and successful adoption across ARSO member states, Africa must move beyond passive attendance and focus on contributing localized data, funding continuous participation for a dedicated pool of technical experts, and driving localized research to actively frame continental problems and needs.
  

• Country Lessons and Adoption Best Practice: The discussion emphasized that successful national AI standardization requires strong sector leadership, long-term government commitment, and strategic funding to sustain expert participation through lengthy multi-year drafting and gazettement processes, as demonstrated by Kenya's two-year journey with the Kenya AI Code of Practice standard. Furthermore, international standards should not simply be adopted blindly; instead, countries must adapt and package high-level international best practices into accessible, localized frameworks, such as step-by-step implementation guides and maturity models, to help local developers and adopters practically measure risk levels and achieve certification.
  

• Opportunities for Improved Coordination and Cooperation: The discussion highlighted that one of the clearest opportunities for regional and international cooperation lies in leveraging the collective diversity of Africa's 55 nations, such as by dividing specialized work paths based on individual country strengths, similar to how Smart Africa tasked South Africa with an AI blueprint, Kenya with a digital economy blueprint, and Tunisia with an innovation ecosystem framework. Regionally, these coordinated standards must connect directly to existing continental frameworks, such as the African Continental Free Trade Area (AfCFTA), where unified standards, such as data maturity models, can securely drive cross-border e-commerce and data interoperability. Internationally, approaches such as a unified template to create a singular, protective framework that global tech providers and the international community can respect are highlighted and encouraged, to be considered alongside a robust, independent pipeline of internal technical capacity, to ensure that foundational standards are determined by African realities rather than by external adaptation. 

  
  

These discussion points contribute to our efforts to broaden awareness, consensus, and foster meaningful engagement and cooperation on AI governance among African stakeholders, to ensure safe, secure, and trustworthy AI.

We thank our panelists for their insightful contributions:  Dr. Victor Odumuyiwa,  Monica Atieno Okoth, and Brian Omwenga Gichana.