Africa expert urges locally designed AI development

A United Nations-backed scientific panel member has urged Africa to develop artificial intelligence (AI) systems tailored to meet local needs rather than importing technology from other regions. Girmaw Abebe Tadesse, head of Microsoft's AI for Good Lab in Kenya, argued that countries should first identify the challenges they want to address before determining the necessary technology, infrastructure, and data.

This perspective was expressed during a presentation in Nairobi, as global leaders and technology executives convened in New York at the UN Headquarters for High-Level Week. Tadesse emphasized that AI policy should begin with a focus on local requirements, stating, "We are trying to bring the reality of our environment and integrate AI technology with the local problems we face."

He highlighted that effective deployment of AI requires more than just hardware; it also necessitates accessible data, skilled operators and users, clear institutional responsibility, cybersecurity measures, sustainable financing, and well-defined routes from priority applications to functioning services. A UN Development Programme assessment revealed that out of 21 AI applications identified across six African deployments, Rwanda identified 42 priority applications across eight sectors.

Tadesse called for collaboration among various stakeholders, including governments, universities, development partners, creative communities, and technology companies, particularly when resources are limited. He stressed that pooling knowledge and expertise can help ensure that AI systems reflect local conditions rather than reproducing assumptions from technologies designed for different markets and societies.

The African Union's Continental Artificial Intelligence Strategy echoes this approach, advocating for an Africa-centered, development-focused, and responsible model of AI. UNDP officials have also highlighted the importance of local data and language representation, given Africa's over 2,000 languages many without significant digital representation.

Regional initiatives are advancing towards greater technological sovereignty, such as the East African Community's commitment to a regional AI technologies fund and the development of systems trained on East African data, capable of supporting regional languages and hosted on regional infrastructure. Tadesse noted that Africa's young population and growing technology ecosystem could become a significant source of AI innovation, with young developers already addressing community problems.

However, he cautioned against fragmented national efforts, urging African countries to collaborate more, share experiences, and learn from each other's successes to avoid repeating the same technical challenges.