An AI Chatbot on WhatsApp Is Coaching Teachers in Northeast Nigeria

aprendIA is an AI powered WhatsApp chatbot from the International Rescue Committee that acts as a personal mentor for teachers in some of the world's hardest classrooms. It matters because Northeast Nigeria is facing an education emergency: 3.5 million people are internally displaced, 10.2 million primary-aged children are out of school, and teachers are often outnumbered 100 to 1. Rather than handing out tablets or building computer labs, the IRC put a teaching assistant inside the one device nearly every teacher already owns: a phone running WhatsApp.

The results so far are striking. As of September 2026, more than 6,000 educators in Borno, Adamawa and Yobe states, known as the BAY states, were using the chatbot, more than ten times the size of the original pilot cohort of roughly 400 teachers. The program expects to give more than 22,000 teachers access by the end of 2026 and aims to reach 137,000 teachers and facilitators by 2028.

A Crisis Inside the Classroom

The global statistics are bleak. According to figures cited by the IRC, there are 234 million crisis-affected school-aged children around the world, and most do not receive the education they need. Fewer than 1 in 5 can read by the end of primary school. UNESCO projects a 44 million teacher deficit by 2030, including 15 million missing teachers in Sub-Saharan Africa alone.

Nigeria illustrates the problem in sharp relief. Only 26 percent of Nigerian children aged 7 to 14 have basic reading and math skills. In the conflict-affected Northeast, where insurgency has displaced millions and destroyed infrastructure, a single teacher can be responsible for 100 students. Sending a child to school is necessary but not sufficient. Quality learning, safe classrooms, and continuity matter just as much, and all three depend on teachers who are supported rather than abandoned.

Help for Today, Skills for the Career

aprendIA, whose name plays on the Spanish and Portuguese word for learning, does two jobs at once. The first is immediate: a teacher facing a large classroom or a tricky topic can ask the chatbot any time and get practical ideas to use in the very next lesson. Classroom management support is the most used feature because it is the help teachers request most.

The second job is long term growth. The bot delivers micro courses that take 5 to 15 minutes to finish, covering Classroom Management, Teaching Literacy, Teaching Numeracy, Teacher Wellbeing, Safeguarding, and Active Learning. Teachers practice a strategy, reflect, and move at their own pace. A teacher named Samuel in Borno State described the change plainly: aprendIA transformed his teaching, his students are learning better, behaving better, and he feels more empowered.

The platform was built with Nigerian teachers, local administrators, and national ministry officials, which is why it fits real classroom conditions. It works on low bandwidth, supports voice to text input, and is fluent in English and Hausa. The same architecture is now being tested in other countries, including Ecuador and Niger, as well as for acute emergency response, with the Airbel Impact Lab designing the education in emergencies version for deployment within two weeks of a crisis.

Why Governments Are Betting on It

Borno, Adamawa and Yobe States have all committed to full rollouts of aprendIA, with written commitments to scale. The reason is largely practical. There is no hardware to procure and minimal training required, because the chatbot runs on phones teachers already carry. Costs fall as reach grows, settling at approximately 13 dollars per teacher per year at scale. The platform is open source, with local capacity building intended for long-term sustainability.

Safety was treated as a design constraint rather than an afterthought. IRC specialists review chatbot responses for quality, accuracy, and safety, and a dedicated safeguarding course equips teachers to protect the whole child. The IRC's broader responsible AI journey shaped this caution. Early internal pilots of AI assistance found systems met quality standards roughly 75 percent of the time and raised staff productivity by nearly 70 percent, but still required human review before responses could be shared safely. That lesson, learned in the Signpost information service that has reached more than 20 million people since 2015, is baked into how aprendIA is governed.

Teachers Actually Keep Using It

Engagement is the quiet metric that separates pilots from impact. In 2025 implementation research, 1 in 3 teachers used aprendIA weekly, and half of those used it 3 to 4 times every week, reporting high satisfaction. For an edtech product in a low-resource setting, that retention is exceptional, and it reflects the decision to build around messaging apps and asynchronous use rather than demanding live connectivity or new devices.

An independent impact evaluation launches in 2027 to measure effects on classroom practice and student learning outcomes in the BAY states. Until then, the evidence is implementation research and adoption curves, which already show more than tenfold growth from pilot to regional scale in a single year.

The Boring Revolution

The most important lesson from aprendIA is counterintuitive. The impactful AI applications in humanitarian settings are not the most sophisticated ones. They are the ones that meet people where they are, on the devices and networks they already have, in the languages they already speak. A WhatsApp chatbot that costs 13 dollars per teacher per year will never make headlines the way a frontier model does. But for a teacher in Maiduguri facing 100 students with no mentor in sight, it is the difference between surviving the school day and growing through it.

The IRC frames its AI philosophy in exactly these terms: AI becomes useful not when it is clever, but when it is boringly reliable. aprendIA is proof that boring reliability, delivered over the humblest channel available, can scale teacher support faster than any training program that came before it.