AI-driven support for bilingual education in Senegal

The challenge

Senegal is implementing an ambitious bilingual education reform known as MOHEBS (Modèle Harmonisé de l’Enseignement Bilingue pour le Sénégal). Under the reform, children learn to read and write in both French and a national language during the early years of primary school.

The reform is based on a simple idea: children learn best when they can build on a language they already know.

Today, MOHEBS is being piloted in schools across 12 of Senegal’s 14 regions. As the reform expands, the Ministry of Education faces a number of implementation challenges; teaching materials must be developed and maintained in multiple national languages. Teachers need training, resources, and ongoing support as they implement new teaching approaches. Providing these resources consistently and at reasonable cost requires new solutions.

The Ministry is exploring whether artificial intelligence could help address some of these challenges.

The opportunity

Rather than asking whether AI can transform education, this project focuses on a more practical question: Can AI help address some of the challenges associated with implementing bilingual education?

To answer this question, we are testing two potential applications of AI: content development and teacher support.

Our approach

IDinsight is working with Senegal’s Ministry of Education and local partners to design, test, and refine AI tools that respond to needs identified by teachers and content developers involved in the reform. Our approach is iterative. We begin by understanding user needs, developing prototypes, testing them with users, collecting feedback, and improving the tools over time.

1. Testing AI for content development

Developing teaching materials in multiple languages requires time and effort. We are exploring whether AI can help content developers produce first drafts of teaching materials more quickly.

The pilot focuses on lesson plans aligned with the national curriculum and existing pedagogical resources. The objective is not to replace content developers, but to support their work and reduce the time required to develop new materials. All generated content is reviewed, revised, and validated before it can be used.

Through this pilot, we hope to better understand where AI can support content development and where human expertise remains essential.

2. Testing AI for teacher support

Teachers implementing MOHEBS receive training before entering the classroom, but many report needing additional support once they begin teaching in national languages.

Teachers may need help understanding specific concepts, finding the right resource, navigating lesson guides, or working with terminology in national languages. Accessing this support can be difficult outside formal training sessions.

To address these challenges, we are developing a teacher support assistant available through WhatsApp.The assistant allows teachers to:

• Access official teaching guides;
• Navigate lesson structures more easily;
• Ask questions about activities and concepts;
• Access support on vocabulary, translation, and pronunciation in national languages.

By making these resources available directly on teachers’ phones, the project aims to provide support that is easy to access when teachers need it most.

What we hope to learn

Across sub-Saharan Africa, around 89% of children cannot read or understand a simple text by age 10 (UNICEF). MOHEBS is a direct response to this crisis. It is now active in pilot schools across 12 of Senegal’s 14 regions, reaching a growing number of children in their early years of schooling.

Our work is still in its early stages. Through these pilots, we are testing whether AI can help address specific challenges faced by teachers and content developers involved in the bilingual reform. We are interested in understanding what is useful, what is not, and how these tools can fit into existing ways of working. The lessons from this work will help guide future improvements to the reform.

If the reform works, it offers a replicable model for how low-resource countries can improve foundational literacy at scale. If it stalls due to funding or capacity gaps, millions of children risk losing access to the quality instruction they need.

Partners

Our work on this project is implemented in close collaboration with the Ministry of Education, ARED, Baamtu, Andakia, and the Gates Foundation.