Education technology has seen unprecedented growth in recent times, accelerated by crises such as Ebola and the COVID-19 pandemic. Amid concerns of increasing inequality in online learning, artificial intelligence has the potential to address disparities and foster inclusive education across Africa by facilitating quality teaching, saving teachers’ time, and enhancing student engagement. 

Education technology (EdTech) continues to advance in Africa, often accelerated by an unlikely factor—health crises. Epidemics such as Ebola and COVID-19 have driven the adoption of innovative approaches in the education sector. These crises have also deepened inequalities in African education systems struggling to grapple with economic and geographical divides. Across Africa, unequal access to EdTech solutions is hampered by the cost of internet services and limited infrastructure. According to a recent EdTech hub report, over three-quarters of education and technology professionals in Africa believe that the move to online learning increases inequality and disadvantages poorer and more marginalized students.  

Meanwhile, in the African context, where recent efforts for universal education have faced capacity constraints, artificial intelligence (AI) can be a crucial ally. It holds the potential to help mitigate quality gaps in an otherwise resource-constrained environment. In 2022, Dalberg supported the conceptualization and design of a collaborative platform that brought together EdTech players working across low and middle-income countries. The team identified gaps in the EdTech ecosystem and conducted a mapping study to identify common areas of interest across 15 EdTech actors, thereby identifying potential areas for collaboration. The team also evaluated existing EdTech networks to assess gaps and pain points in collaboration. EdTech Collab uses AI to curate and recommend relevant content, resources, and opportunities for its users based on their profiles, interests, and goals. The platform also leverages AI to generate insights and evidence on the impact and best practices of EdTech interventions in the South African context. 

“A number of cost-effective innovative solutions hold promise in elevating the quality of education, but these solutions are often absent in public schools, which educate more than 80% of school-going children across the continent.” 

– Unlocking Public Finance for Education Innovations with Elimu-Soko: Lessons from Rwanda 

As the AI revolution enters classrooms, proponents of AI argue that the technology can 

• Facilitate quality teaching by supporting pedagogy improvement, which can be particularly impactful in classrooms with high student-to-teacher ratios, ensuring no student is left behind. 
• Save teacher time by reducing the time taken to complete non-core tasks, helping to dedicate more time to direct student support. 
• Enhance student engagement through personalized intelligent tutoring and peer-assisted learning (PAL), providing individualized assistance aimed at improving learning outcomes and reaching students not enrolled in traditional schools, thus expanding educational access.

Roadmap to Implementation

As debates surrounding the role of AI in education persist, school networks on the ground are actively working and investing in new models with the hope of empowering their teachers and students.  

Rising Academy Network (“Rising”) is a fast-growing education social enterprise that supports the quality of education in both private and public schools. Today, Rising is working with over 900 schools supporting the education of over 250,000 students across Rwanda, Sierra Leone, Liberia, and Ghana. Teachers at Rising schools undergo extensive training on delivering quality lessons and adhering to a rigorous curriculum. Independent evaluations have found that Rising students make learning gains two to three times greater than peers in comparable schools. The Rising network collects extensive data on effective practices, sharing results with governments and other educational organizations to improve learning for students across all schools. 

Rising is exploring a multifaceted approach to incorporating AI in their classrooms, aiming to overcome challenges posed by technology while maximizing its impact on student learning outcomes. At the core of their approach is the desire to streamline administrative tasks and enhance the overall teaching and learning experience, leveraging AI support for teachers without replacing them.  

In this process, Rising is exploring different approaches to leverage AI in their work:  

1. Facilitate quality teaching by supporting teachers’ preparedness for content delivery: Recognizing the importance of well-prepared teachers, Rising is exploring the use of large language models (LLMs) to create interactive prompts for teachers in advance of lessons. These prompts aim to reinforce teachers’ understanding of lesson content before delivery and provide key reminders and practice opportunities. In the future, school leaders could receive alerts about teacher preparedness, allowing for additional school-based support to those teachers. While this strategy is in the early stages, it can enable teacher improvement and effective classroom instruction.
2. Enabling timesaving for teachers through automatic grading: Grading handwritten assessments consumes a significant amount of teacher time. To reduce this, Rising is experimenting with optical character recognition (OCR) but initial results showed only 62% accuracy. Undeterred, Rising Academies explored alternative tools, eventually finding an 86% accuracy rate, with Amazon Text Tract. Rising is committed to attaining a high accuracy rate to ensure that students’ exams are properly graded. This remains a promising solution that, despite the need for some manual post-processing, is expected to significantly reduce the burden on teachers and provide valuable insights from granular assessment data. 
3. Enhancing student engagement by making math tutoring fun and accessible: Rising Academies developed Rory, a WhatsApp-based math tutor that combines a low-tech front end, a high-tech back end (natural language processing (NLP), Generative AI), with a representative data set (African learners). Rori is designed to reach students who do not have access to laptops or high-bandwidth internet connections. Initial attempts to incorporate LLMs for math tutoring faced accuracy issues, prompting Rising to insert a middle layer between the student and the generative AI engine. Rising has recently completed an extensive red-teaming/safety testing process. This has allowed for more focused and context-specific interactions, enhancing the learning experience. Going forward, Rising Academies will continue to ensure that their AI-math tutor is responsive to students’ performance namely by developing generative chats that help students set, track, and reflect on learning goals and making the content more accessible to students whose native language is not English and struggling readers by incorporating text-to-speech functionality.

On the Brink of Change

Rising’s journey illustrates both the potential of AI as well as the constraints and risks that come with it. Their experience has shown the importance of moving beyond algorithms to explore strategies that prioritize ongoing enhancement, safeguard grading accuracy, and foster respectful and responsive interactions with education practitioners—in particular, teachers. The use of AI technology in support of teachers offers a promising solution, potentially serving as a complement to their work, enhancing teaching effectiveness and tailoring approaches to diverse student needs. While AI presents a meaningful opportunity for automation and scalability, democratizing AI requires a commitment to maintaining high standards of quality and adaptability to evolving educational needs.