Google has partnered with leading African research institutions to launch WAXAL, a speech dataset aimed at expanding access to artificial intelligence for over 100 million people across Sub-Saharan Africa.
Developed three years with Google’s support, WAXAL contains 1,250 hours of transcribed natural speech and over 20 hours of studio-quality recordings. These resources are expected to support the development of voice recognition systems and synthetic speech tools for African users.
The initiative targets a major digital gap by providing foundational speech data for 21 African languages. These include spoken languages such as Hausa and Yoruba, as well as Luganda and Acholi.
Despite the rapid global growth of voice-enabled technologies, many African languages remain underserved due to limited high-quality speech datasets. This situation has left millions unable to access digital tools in their native tongues.
Aisha Walcott-Bryant, head of Google Research Africa, said the project is designed to empower African students, researchers, and entrepreneurs. She said it will help them build technologies in local languages and unlock economic opportunities across the continent.
Data collection for WAXAL was led by African institutions, including Makerere University in Uganda, the University of Ghana, and Digital Umuganda in Rwanda. The institutions retain ownership of the dataset.
