Building Frontier Education Organizations With Artificial Intelligence and Agentic Systems in East Africa: A Systematic Review
This systematic review of 28 sources identifies critical infrastructure, capacity, and governance barriers to AI adoption in East African education while highlighting emerging local innovations and proposing a four-pillar framework to guide the development of sustainable, context-sensitive frontier education organizations.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Imagine a classroom where the teacher does not just answer questions but actively plans the day's lessons, remembers what each student struggled with yesterday, and independently finds the right tools to help them learn tomorrow. This is the promise of a new kind of artificial intelligence, often called "agentic" systems. Unlike the chatbots that simply wait for a user to type a command, these systems can set their own goals, take a series of steps to reach them, and keep a running memory of their progress. For a region like East Africa, where schools often face severe shortages of teachers and limited access to electricity or the internet, this technology offers a potential lifeline. It could theoretically provide personalized tutoring to millions of students who currently have no one to guide them. However, the question remains whether these advanced tools can actually work in places where the basic infrastructure is still being built, and whether they will help everyone or just widen the gap between the rich and the poor.
A team of researchers from Aksum University in Ethiopia set out to answer these questions by looking at the entire landscape of artificial intelligence in education across East Africa. They conducted a systematic review, which is a rigorous method of gathering every available piece of evidence on a topic to see the full picture. They searched through thousands of reports, academic papers, and government documents from 2020 to 2026, covering eight partner states of the East African Community as well as Ethiopia. Their goal was to move beyond the hype and find out what is actually happening on the ground: what is being built, what is working, and what is stopping these organizations from becoming truly "frontier" institutions—places that use technology to deeply transform learning rather than just adding gadgets to old systems.
The researchers found that while many countries in the region have created ambitious national policies and strategies for artificial intelligence, the reality inside the schools and universities tells a different story. The most significant barrier is not a lack of ideas, but a lack of basic readiness. The review highlighted a severe shortage of qualified teachers; across the universities in the region, there is an average gap of about 50 percent in the number of lecturers needed to meet student demand. In Kenya alone, the shortage is projected to grow to more than 35,000 lecturers by 2030. This creates a dangerous temptation to view artificial intelligence as a substitute for hiring human staff, but the researchers argue this would be a mistake. Without enough trained people to guide the technology, the tools cannot function effectively, and the risk of misuse grows.
Beyond the human shortage, the physical environment poses a massive challenge. The internet infrastructure in the region is fragile and often insufficient for running complex AI systems that require constant online connections. In some areas, traffic between neighboring capitals is so poor that data has to travel all the way to Europe and back just to connect two African cities. This reality has forced innovators to get creative. Instead of waiting for perfect internet, some groups have built systems that work offline on standard laptops, while others have designed tutoring tools that run entirely on basic text messages sent to old-fashioned mobile phones. These "offline-first" approaches are the most promising examples of frontier education in the region, showing that technology can be adapted to local constraints rather than forcing the local environment to change to fit the technology.
Despite these clever adaptations, the review uncovered a critical gap in the evidence: no one knows for sure if these tools are actually improving learning outcomes. The researchers found dozens of reports describing new projects, such as a WhatsApp-based assistant in Tanzania or a national teacher training program in Rwanda, but very few of these projects included independent, scientific tests to prove they worked. Most of the success stories come from the companies that built the tools or the organizations that launched them, rather than from neutral researchers. In fact, the only rigorous study that proved an AI tutor improved math scores in Africa was conducted in Sierra Leone, outside the specific region this review focused on. This lack of solid proof means that governments and schools are investing in technology without knowing if it is truly helping students learn.
The researchers also pointed out that the rules governing these technologies are not yet strong enough to protect students. In many developed regions, laws require strict testing for bias and human oversight before AI can be used in schools. In East Africa, most policies are still just suggestions or guidelines. This leaves students vulnerable to data privacy risks and decisions made by machines that no one is watching. The review emphasizes that for these education organizations to truly succeed, they must shift their focus from simply buying technology to building the capacity of their people. This means training teachers to use these tools critically, ensuring that humans remain in charge of important decisions, and investing in infrastructure that reaches the most remote and underserved communities.
Ultimately, the study concludes that building a frontier education organization is not about having the most advanced AI, but about having the right foundation to use it well. The path forward requires a coordinated effort across the entire region to share data, train teachers, and create laws that protect students. It demands that leaders stop treating artificial intelligence as a magic solution to staffing shortages and start treating it as a tool that needs careful management. If East Africa can build this foundation, focusing on human capability and regional cooperation, it has the potential to create a model of education that is not just catching up to the rest of the world, but leading the way in how technology can serve humanity in diverse and challenging environments.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.