What Two Education Summits Said About AI — and the Part Worth Holding Onto

Two education summits in Africa this month arrived at the same place, independently.

At NESS 2026 in Lagos on 18 July, a school-leader session on AI closed with this line: "AI will enable those with clarity of vision." At Kenya's Freedom 250 EdTech Summit in Nairobi, a full-day digital learning track concluded that outcomes depend on how well institutions use digital tools — not on how many they have.

Neither summit was being sceptical about technology. Both were describing a sequence: clarity first, tools second.

Most of what Ghanaian school proprietors are being told about technology describes a different sequence. Buy the system, and visibility will follow. Get the software and you'll know what's happening in your school. The technology will surface the insights.

It won't. Not without something that comes before it.

The question that precedes every system

The schools that get something useful out of technology have usually answered one question before they signed anything: what, specifically, do I need to see that I can't see today?

Not what the demo said the system could do. What real thing — a number, a trend, a pattern — do you reach for and find missing? A proprietor who can name it clearly has already done most of the hard work. The tool just needs to surface it.

Three questions worth sitting with before adopting any system:

1. What do I need to know that I don't know today? Vague answers — "I want more visibility," "I need to know if things are on track" — lead to vague purchases. Something specific is useful: "I want to know, by week three of term, which families are likely to carry an arrears balance into the next term." That's a design requirement. It can be built toward. A vague feeling cannot.

2. How early do I need to know it? Most problems in schools are not sudden. Fee arrears build slowly. Attendance drift is visible weeks before it shows up in performance. A staffing gap appears in the timetable before it appears in a resignation. The value of good operational information is rarely the data — it's the timing. "Early enough to act" is the design requirement. Build toward that.

3. What would I do differently if I knew it? This is the honest check. If the answer is "probably nothing — I know what to do, I just don't have time or resources," the problem is not information. Fix the time or resources first. If the answer is "I'd have that conversation a month earlier" or "I'd adjust the fee structure for that intake cohort," you have found the signal worth surfacing. A system that surfaces it will earn its keep.

What the Lagos session was really saying

"AI will enable those with clarity of vision" is not a compliment to AI. It is a description of a dependency. A system given to a school without a clear sense of what it needs to see will produce overhead, not insight — more screens, more data, more time managing the tool rather than the school.

The schools across Nigeria and Kenya that the summits described as succeeding with digital tools were not the ones with the newest software. They were the ones who had already developed a clear operational picture of their schools — and then found systems that made that picture easier to maintain.

That sequence is available to every school leader in Ghana, starting with a notebook and three honest answers to the questions above. The AI comes after.

As the operational picture of a school grows more complex — more students, more staff, more payment channels, more terms of data — maintaining it manually becomes the bottleneck. That is when structured systems stop being a nice-to-have. The clarity still comes first. The system serves it.