Per our reporting

African leaders recently urged for a greater say in shaping global AI standards, amid growing concerns over AI misuse and potential rogue behavior. At a U.N. Security Council meeting, Liberia's permanent representative Lewis Garseedah Brown II emphasized that Africa must play an equal role in determining AI ethics and architectures.

African nations are racing to deploy AI systems developed in the U.S. and China, but many lack the means to independently test these systems for safety, experts warn. Jonathan Shock, a University of Cape Town professor, highlighted the lack of discussion on AI safety, creating a significant void. Fragmented regulations could complicate global deployments and increase costs, posing risks of AI misuse, such as Kenya's health insurance algorithm driving up costs for the poor and West African scammers using AI-generated content for extortion.

Militant groups like Boko Haram have also leveraged AI to plan attacks and design weapons. Out of 85 large companies surveyed for a PricewaterhouseCoopers report, over 80% were running AI pilots, yet many lacked formal safety oversight structures. Companies face trade-offs between American and Chinese models, each with unique risks, and the ability to shut down models remotely varies significantly.

African governments are starting to incorporate AI in sensitive sectors, but fewer than half have AI policies. Kenya is the only African country participating in an international AI safety network, which recently tested AI agents for risks such as data leaks and fraud. However, these tests focused on common global issues rather than African-specific concerns.

African nations need to develop their own safety evaluation capacities and gain genuine representation in multilateral processes to shape AI technologies that impact their economies and populations uniquely.