Integration of Effective, Inclusive and Equitable Learning Innovation in Basic Schools: The Effect of Teacher-Facing AI Support on Teaching Effectiveness and Student Learning

Despite substantial improvement in access, Ghana continue to face persistent challenges in improving learning outcomes, particularly in large, heterogeneous classrooms, partly due to low active engagement, limited resources, limited instructional time and wide variation in students’ learning needs. This project examines whether early introduction teacher-facing AI improves teaching effectiveness, addresses learning variability within classrooms and across socioeconomic groups, and improves student cognitive and non-cognitive outcomes. The project will be piloted in 40 schools in the Greater Accra region within a year. The project will provide teachers in treatment schools with training and structured support to use Khanmigo AI as an instructional support tool. Teachers will use the tool for lesson planning, clearer explanations with real-world examples, designing differentiated instruction and formative assessments for low-and high-performing students. Control schools are untreated. This study moves beyond the average effect to the examination of learning variability, assessing whether AI helps narrow achievement gaps and improve learning trajectories for lower-performing students without disadvantaging higher performers. The project is implemented in partnership with the Ghana Education Service. In a resource- constrained setting, this free AI tool and intervention approach provide a low-cost, scalable way to improve instructional quality and learning equity without massive infrastructure investment.