From Chalkboards to Chatbots: How Generative AI is Reshaping Classroom Learning in Nigeria

Nigeria is facing a severe educational challenge driven by a massive demographic surge where the median age is just 18 years. According to a UNESCO report on educational technology in Africa, sub-Saharan Africa will require 15 million new teachers by 2030 to meet demand. In Nigeria, where public university lecture halls can hold up to 600 students and private tutors charge between 3,000 and 6,000 Naira per hour, educational stakeholders are turning to artificial intelligence. Recent trials and academic studies across the country are showing that when properly structured, generative artificial intelligence can act as an effective, scalable virtual tutor.

The Edo State Generative AI Pilot Program

A landmark pilot study conducted in Edo State, Nigeria, has provided empirical proof of how generative artificial intelligence can boost learning outcomes. The six-week after-school program, which ran between June and July 2024, targeted first-year senior secondary students across nine public secondary schools, including Edo Boys High School in Benin City. Evaluated through a randomized controlled trial by World Bank and Stanford University researchers, including Martin De Simone, Federico Tiberti, Maria Barron Rodriguez, Federico Manolio, Wuraola Mosuro, and Eliot Jolomi Dikoru, the intervention yielded highly positive results.

During the program, students worked in pairs under direct teacher supervision using Microsoft Copilot, which is powered by GPT-4. Rather than using the technology to bypass thinking, the students used prompts specifically designed to encourage reasoning. The results of a pen-and-paper assessment taken after the program showed a significant overall improvement of 0.31 standard deviations across English language, artificial intelligence knowledge, and digital skills. In English language alone, which was the main focus of the pilot, students achieved an improvement of 0.23 standard deviations. Researchers noted that these gains are equivalent to 1.5 to 2 years of traditional schooling, ranking the intervention among the most cost-effective learning programs available. The evaluation also showed that while all students benefited, the largest positive effects were recorded among female students and those with higher baseline academic performance.

The Risk of Over-Reliance and the University of Port Harcourt Study

While the Edo State pilot highlights the potential of structured artificial intelligence integration, experts warn against the risks of cognitive offloading, where students use chatbots as a substitute for active thinking. This contrast is reflected in a separate study conducted by Stella Eteng-Uket and Ebere Ezeoguine at the University of Port Harcourt. Using a quasi-experimental design, the researchers analyzed the impact of artificial intelligence chatbots on higher education students, dividing 186 participants into an experimental group of 90 students and a control group of 96 students.

The University of Port Harcourt study found no significant difference in learning outcomes or student engagement between the group that used chatbots and the group that did not. The researchers concluded that simply providing access to conversational agents is not enough, emphasizing the need for a balanced integration of artificial intelligence with traditional teaching methods to achieve optimal educational results.

Scalable Solutions for Unified Exams

Despite differing academic results, the demand for digital learning companions in Nigeria continues to surge. Over 1.9 million candidates register annually for the Joint Admissions and Matriculation Board Unified Tertiary Matriculation Examination. To prepare for this highly competitive gateway to tertiary education, students are increasingly adopting digital platforms. Developers are building WhatsApp-based chatbots and virtual learning tools fine-tuned specifically on the Nigerian curriculum, including West African Examinations Council and Joint Admissions and Matriculation Board preparatory materials. These tools operate around the clock, offering localized explanations in English as well as Hausa, Yoruba, and Igbo to help bridge the learning gap for rural and underserved students.

What this means for Africa: Carefully designed generative artificial intelligence tools, when paired with teacher supervision and structured reasoning prompts, offer a highly cost-effective method to combat severe teacher shortages and improve secondary student learning outcomes across the continent.