African nations are demanding a greater role in establishing global standards for artificial intelligence. During a recent United Nations Security Council meeting, representatives argued that African countries must act as co-architects in defining the ethics and architecture of AI to avoid being mere consumers or testing grounds for foreign technology.
The region faces significant challenges as it adopts AI systems developed in the U.S. and China. According to a PricewaterhouseCoopers survey, over 80% of 85 large companies are running AI pilots, yet many lack formal safety oversight. Experts note that fewer than half of African countries currently have established AI policies or national strategies.
Local incidents highlight the risks, including algorithmic bias in Kenyan health insurance and the use of AI by militant groups for weapon design. Currently, only Kenya participates in an international AI safety network. Researchers argue that existing safety evaluations often overlook regional contexts, such as the specific economic impact of AI-related harms in African nations.
Experts emphasize that the ability to control AI models is critical. While closed-weight models like ChatGPT are centrally managed, open-weight alternatives such as DeepSeek and Kimi can be modified, making remote shutdown difficult. Analysts suggest that the focus should shift toward who holds the off-switch rather than solely the country of origin.
