
A maths tutor in Ghana named Rori recently posed a question to a student: “What is the total number of students who like football?” After the student analyzed a bar chart and gave an incorrect answer, the feedback came not from a human instructor but from an automated assistant over WhatsApp. This exchange marked an early experiment in how artificial intelligence might transform education across Africa.
Initial findings indicate AI could accelerate learning outcomes by months or even years. In Ghana, a study involving 500 students showed that those using Rori—a chatbot developed by Rising Academies, which partners with 1,400 government schools—achieved the equivalent of an extra year’s learning in just eight months compared to a control group. The system goes beyond simple answers by analyzing responses, providing hints, and encouraging independent problem-solving.
Separate trials in Sierra Leone used Google’s Gemini chatbot with comparable results. Students engaging with the Guided Learning feature matched the progress of 1.2 to 1.7 years of traditional instruction within eight weeks. Unlike generic chatbots that might offer quick fixes, Gemini employed teaching techniques such as scaffolding, breaking complex problems into smaller steps, to replicate effective instruction.
Mathias Esmann, director of partnerships at Fab AI, which assisted in the Sierra Leone study, cautions that unstructured AI access can harm learning. “If you just give students access to a chatbot, it could actually have detrimental effects, because it just becomes a tool for finishing your homework really quickly,” he says. However, when designed to guide learning, such as through scaffolding techniques, AI can produce meaningful educational gains.
Infrastructure and policy barriers slow AI adoption
The main obstacle remains scaling these tools. Most pilot programs have focused on urban areas where smartphones and electricity are more dependable, yet fewer than one-third of Africans have mobile internet access. In Sierra Leone’s trial, researchers provided devices and field monitors to maintain charging, a level of support absent in typical classrooms.
Admire Mare, head of the Department of Communication and Media Studies at the University of Johannesburg, highlights that AI’s potential in Africa is still constrained by infrastructure and policy gaps. “Certainly we have seen an uptick of the use of AI, but of course most of it is not really guided by any institutional policies,” he notes.
Several startups are already adapting AI to local needs. In Mali, RobotsMali uses ChatGPT and DALL-E to create learning materials in Bambara. In Kenya, Kytabu is developing AI-powered smart glasses for teachers, offering real-time lesson suggestions and student engagement data. Meanwhile, M-Shule sends AI-generated quizzes and summaries via WhatsApp or SMS, designed for low-bandwidth environments.
Startups tailor AI tools to Africa’s unique needs
Teachers emphasize that human interaction remains central to learning. “Rori interprets what they wrote and gives feedback, and if they’re stuck, they get a hint first and a chance to try again,” explains Hannah Horne-Robinson, Rori research and assessment manager.
Rising Academies created the Rori chatbot after the pandemic forced schools to rely on radio and text-based learning. Math specialists design each question, while the AI evaluates responses and provides hints before revealing solutions. The platform delivers content as plain text, ensuring compatibility with basic phones common in many classrooms.
During Sierra Leone’s trial, educators observed student interactions to ensure discussions focused on concept development rather than quick answers. RobotsMali combines ChatGPT and DALL-E to produce Bambara-language reading materials, expanding locally relevant resources. The Edutech Institute imports Indian educational software, adapting it for African curricula through AI customization. M-Shule automates quiz creation and sends summaries via WhatsApp or SMS, optimized for low-bandwidth conditions.