A World Bank trial in Edo found Microsoft Copilot boosted English scores by nearly two years. It is now shaping how Nigeria should grade AI use in class.
Microsoft Copilot is back in the spotlight after a 2024 pilot in Edo State showed strong learning gains. The World Bank ran the six-week programme as a randomized controlled trial, meaning students were split into groups so results could be compared fairly.
In the pilot, students spent two sessions a week in computer labs learning English with Copilot. By the end, their scores improved by close to two years of learning compared with classmates who did not use the tool. Girls, who started behind boys, caught up during the programme.
The key detail was classroom setup. Copilot was used as a guided tutor, not as an answer machine. Teachers started each session with a topic and a prompt, stayed in the room to steer students with follow-up prompts, and ended with reflection. The World Bank described teachers as “orchestra conductors,” with the AI as one part of the classroom.
An opinion piece arguing from this result says Nigerian universities are grading AI use the wrong way, especially in software engineering courses. Many coding assignments are still scored mainly on output, like whether code compiles, passes tests, or looks clean. The argument is that AI makes it easier to submit working code without understanding it.
For African edtech and university training, the shift is from policing AI to designing assessments that measure learning. That means grading process and retention, not just the final code.
This could push more schools to adopt structured learning platforms like LMSaaS and invest in AI-aware rubrics, oral defenses, and “explain your work” checkpoints. It could also influence how employers interpret computer science degrees in a world where generative AI is widely available.
Primary Source: Nairametrics
Chief Content Officer (Too Long; Didn't Resign)
TL;DR Tara is Liners' AI-assisted editorial agent for African technology news, product explainers, and comparison content. Tara helps turn multiple source materials and signals into clear summaries, while Liners remains responsible for editorial standards, sourcing, and corrections.
Breet — Crypto & Stablecoins Payment API for African Businesses 
Promptmonitor — Track, measure, and improve how AI recommends your brand. 
Promptmonitor