Two African students studying in the United States have won a prestigious global award for an artificial intelligence education system that works entirely without internet access, using only basic mobile phones.

Happy Niyorurema, from Rwanda, and Senegalese student Mame Niang, both at Texas Christian University, developed a platform that delivers AI-powered learning through ordinary voice calls. Their project was honoured with the Global Best M-Gov Award in Dubai, a prize that recognises breakthrough mobile government solutions from around the world.

The pair set out to tackle a problem that is often overlooked in the rush to adopt digital tools: billions of people, especially in the Global South, remain offline. Instead of assuming access to smartphones, data bundles or stable networks, they built an AI system that runs on telephony infrastructure, allowing any learner with even the simplest handset to call in and receive tailored educational support.

Behind the scenes, their custom large language model processes questions and delivers explanations in real time, much like popular online chatbots, but without requiring a data connection. Lessons, quizzes and explanations are converted into interactive voice responses, so a student can dial a number, follow prompts and ask questions in a conversational way.

The system is already being piloted in Rwanda, where many households own basic mobile phones but lack reliable internet. Early users include secondary school pupils preparing for exams and adults seeking foundational skills. Niyorurema and Niang say they are working with local educators to align content with national curricula and to support multiple languages so that learners can study in the tongues they use at home and in class.

Plans are under way to expand the service to other African countries, starting with Senegal and Zambia. The team is exploring partnerships with governments, telecom operators and education ministries to scale the platform at low cost, arguing that existing voice networks can be repurposed as powerful learning channels.

The recognition in Dubai places their work among leading examples of how artificial intelligence can be adapted to real-world constraints rather than deepening digital divides. For Niyorurema and Niang, the goal is simple but ambitious: to ensure that a child in a remote village with only a basic phone can access the same quality of explanations and practice as a student in a well-connected city classroom.