• Uses small language models suited for low-data, low-connectivity settings
  
• Aims to localize AI for East Africa and close language access gap
  

Microsoft has unveiled Project Gecko in Kenya, an initiative designed to tackle one of artificial intelligence’s biggest gaps: the lack of support for African languages.

Launched in Nairobi, the program aims to make AI more accessible to people in Kenya by developing models that can understand and generate local languages such as Swahili, Kikuyu and Dholuo.

Unlike large AI models that are primarily trained on languages dominating online content, Project Gecko uses lighter, modular “small language models.” These systems are built to perform well in environments with limited data and inconsistent connectivity, and to power digital assistants tailored to East Africa’s linguistic and cultural context.

“Gecko seeks to reverse this dynamic — building AI systems from the ground up, shaped by the knowledge, languages, and modalities of the global majority. Achieving population-scale impact requires a fundamental rethinking of how AI is localized, evaluated, and deployed,” Microsoft said in a statement.

The initiative is part of Microsoft’s broader effort to expand AI access across what it calls the “global majority,” regions that remain underrepresented in most existing models. By developing tools that work beyond major world languages, the company aims to help close the language gap that limits AI adoption across Africa.

Adoni Conrad Quenum