Togo has launched a national initiative to develop artificial intelligence models for all 50 of its national languages, marking a significant step toward making AI more inclusive and accessible.
Announced at the AI for Good Global Summit in Geneva, the initiative brings together Togo AI Lab, Zindi and Umbaji to build open-source datasets and AI models capable of supporting speech recognition, text-to-speech and machine translation in local languages.
The partners aim to collect at least 50 hours of validated speech and 6,000 translated sentence pairs for each language, with the resulting datasets powering four open AI competitions on Zindi’s platform. The competitions will offer a combined $40,000 in prize funding to accelerate the development of African language AI models.
The initiative aligns with Togo’s National AI Strategy and comes amid growing global efforts to address the underrepresentation of African languages in AI systems. According to UNESCO, more than 2,000 African languages remain largely absent from the datasets used to train today’s AI models.
