The Lower Eastern Times Opening The Third Eye
Related Articles
Google has expanded its WAXAL speech dataset to include Kikuyu, Dholuo, and Luganda, a move aimed at speeding up the development of artificial intelligence tools for millions of African language speakers who have historically been left out of voice-based technologies.
The announcement was made during a launch event in Nairobi on Monday, February 2, 2026. The dataset is intended to support developers and researchers in building AI systems that can understand African languages, helping to close a long-standing gap in access to digital services across the continent.
By incorporating Kikuyu, Dholuo, and Luganda, Google seeks to enhance access to AI-powered applications in East Africa, such as voice assistants, speech-to-text tools, learning platforms, and digital public services.
Google said the dataset contains 1,250 hours of transcribed natural speech, along with more than 20 hours of high-quality studio recordings collected over a three-year period.
Aisha Walcott-Bryantt, Head of Google Research Africa, said the broader goal of WAXAL is to empower African communities. She noted that in areas with limited English proficiency, the technology could significantly improve sectors like education, agriculture, and healthcare by enabling services in local languages.
She added that the dataset offers a vital foundation for students, researchers, and entrepreneurs to create technology in their own languages, potentially reaching more than 100 million people.
Data collection was carried out by African academic and community institutions—including Makerere University in Uganda, the University of Ghana, and Digital Umuganda in Rwanda—working alongside Google researchers. In total, data was gathered for 21 local languages across 25 countries.
In addition to newly added languages, the dataset also includes Swahili, which is widely spoken in Kenya and beyond.
The WAXAL speech dataset is released under a Creative Commons license, giving developers broad flexibility to use it within African contexts.
Overall, the expanded dataset supports the development of technology that understands local African languages, helping to make digital tools more inclusive and reducing the technology gap affecting over 100 million people across the continent.
