Lost in translation: How Africa is trying to close the AI language gap

A new initiative in Africa is tackling the AI language gap by releasing the largest dataset of African languages to date, funded by the Gates Foundation, to ensure that technological advancements are accessible to all speakers on the continent. Africa boasts a rich linguistic diversity with over 2,000 languages, but AI tools like ChatGPT are predominantly trained on English, European, and Chinese languages due to the abundance of written text. This has left many African languages, often oral, underrepresented, creating a digital divide that excludes millions from AI benefits. The Africa Next Voices project, with a $2.2 million grant, has compiled AI-ready datasets for 18 languages, including Hausa, Yoruba, isiZulu, and others. Over two years, researchers recorded 9,000 hours of speech in Kenya, Nigeria, and South Africa, capturing real-world scenarios in farming, health, and education to make the data practical and inclusive. This open-access data enables developers to build tools that translate, transcribe, and respond in local languages. For example, the AI-Farmer app helps South African farmers like Kelebogile Mosime by providing agricultural advice in Setswana, significantly aiding those in rural areas with limited exposure to technology. Experts involved stress the critical need for such initiatives. Prof Vukosi Marivate from the University of Pretoria emphasized, ‘We think in our own languages… If technology doesn’t reflect that, a whole group risks being left behind.’ Computational linguist Lilian Wanzare added that gathering diverse voices ensures the data reflects actual usage, something big tech often overlooks. Beyond functionality, there’s a profound cultural impact. Pelonomi Moiloa, CEO of Lelapa AI, pointed out that language barriers can prevent access to essential services, while Prof Marivate warned that failing to include indigenous languages could lead to the loss of cultural heritage and knowledge. This project not only bridges the technology gap but also safeguards Africa’s linguistic diversity for future generations, with plans to expand to more languages and promote equitable AI development globally.