In a major step toward inclusive technology, Google has partnered with African universities and research groups to launch WAXAL, a project designed to train artificial intelligence systems in African languages.
WAXAL is a free, large-scale speech dataset created to support research and make AI tools more accessible to African language speakers. It aims to close a digital gap affecting over 100 million people whose languages are often excluded from voice-based technologies like virtual assistants and speech-to-text tools.
Developed over three years and funded by Google, the dataset includes 1,250 hours of transcribed everyday speech and over 20 hours of studio-quality recordings for creating natural-sounding artificial voices.
Head of Google Research Africa, Aisha Walcott-Bryant, said the initiative will empower students, researchers, and entrepreneurs to build technology in local languages, opening new opportunities in education, voice services, and innovation.
A key feature of WAXAL is its African-led approach. Data was collected by institutions including Makerere University in Uganda, the University of Ghana, and Digital Umuganda in Rwanda, with African partners retaining full ownership of the data.
Despite Africa having more than 2,000 languages, most AI systems support only a few global ones. WAXAL currently covers 21 African languages, including Hausa, Yoruba, Igbo, Luganda, Swahili, and others, marking a significant step toward inclusive artificial intelligence.
About the Author
