Prime Minister Dr. Justin Nsengiyumva of Rwanda addressed the United Nations General Assembly, highlighting concerns about the governance of artificial intelligence (AI). He emphasized that countries that do not develop frontier AI models should not blindly trust systems developed elsewhere, given differences in populations and conditions. Nsengiyumva stressed the need for governments to assess the safety, reliability, and accountability of AI systems before deploying them.
Nsengiyumva described a global conversation on AI governance that is effectively taking place in two rooms. In one, leading AI companies discuss standards and safety among themselves. In another, governments of countries competing over the technology negotiate with one another. He warned that this gap is becoming harder to ignore as governments seek ways to cooperate on AI safety. Countries are entering the AI era from different starting points, with some integrating AI into public services and workplaces, while others are still building the necessary infrastructure.
President Alar Karis of Estonia, who also addressed the UN gathering, emphasized that the choice should not be between developing AI quickly and making it safe. He advocated for common methods for evaluating AI systems and better cooperation to prevent serious incidents that can cross borders. This adds another layer to Rwanda's argument, where the issue is not just about accessing AI, but also about having the knowledge and institutional capacity to understand what is being provided.
The divide is not simply between countries that have AI and those that do not. It is increasingly between countries that can turn access into useful, locally relevant applications and those that remain dependent on systems they have limited capacity to examine or adapt. This distinction is becoming crucial as AI moves into areas where mistakes can directly affect people. Africa does not want to remain just a customer of AI built elsewhere.
Kenya's Special Envoy for Technology, Philip Thigo, pushed the argument further, stating that Africa cannot afford to be a continent that only consumes AI built elsewhere. He emphasized the need for technical and institutional capacity to make meaningful choices about technologies being deployed on the continent. Canada's Ambassador to the UN, David Lametti, approached the problem from the perspective of responsibility, stressing that developers of powerful AI systems should carry responsibility when things go wrong.
Rwanda's National AI Agency is ready to serve as a regional host for shared AI testing, providing technical expertise to help countries across Africa build the ability to independently evaluate advanced systems. Nsengiyumva warned that most states cannot assess a frontier model today and will not be able to for years, making inclusion without capacity ceremonial. The goal is to turn Rwanda's National AI Agency into part of the infrastructure of AI safety in Africa.
The challenge is becoming more immediate as AI systems grow more autonomous and their failures can cross organizational and national borders. A UN scientific panel noted that no single company or country is likely to see enough incidents on its own to identify every emerging pattern. Nsengiyumva concluded that AI governance cannot be inclusive if only those building the technology have the capacity to judge it.
Key points
- Countries that do not develop frontier AI models should have the capacity to test and evaluate AI systems.
- Rwanda's National AI Agency is ready to serve as a regional host for shared AI testing.
- AI governance should be inclusive, with countries that do not develop AI having a say in its development and deployment.
