The four stages every employee climbs, and how to measure them.
Literacy, Fluency, Mastery, Stewardship, made concrete and measurable.
"Upskill the workforce" is the most common line in any AI strategy, and the least useful, because it cannot be measured. You cannot manage what you cannot see. So we describe AI capability as four stages every person climbs, each with a test you can actually apply.
Literacy. The person understands what AI can and cannot do, and how to use it safely. They can write a clear instruction, refine it when the first answer is weak, and judge whether the output is good or wrong. The test: can they take a real task, get a useful result, and tell you why they trust it? Most people, in most South African businesses, start here. That is not a weakness. It is the right place to begin.
Fluency. The person uses AI regularly to improve the speed and quality of their own work. They have built small shortcuts: a saved prompt, a reusable template, a checklist that works. The test: has their everyday work changed, and can they show you a task that now takes less time because of something they made?
Mastery. The person builds repeatable workflows, connects tools, and solves real problems in their role. They are no longer using AI so much as designing with it. The test: have they built something other people now rely on?
Stewardship. The person lifts the people around them. They set the standard, spot where AI helps and where it does not, and carry the rest of the team up the ladder. The test: is the team better because this person is on it? Stewards are rare, and they are what makes a programme self-sustaining after the consultants leave.
Why bother measuring it
Because skill, not technology, is the constraint. When South Africa's financial regulators surveyed the sector in late 2025, the single biggest barrier to AI adoption was a shortage of talent and skills, ahead of data quality, explainability, and cost. You cannot close a gap you have not measured. Place each person on the ladder, and three things become possible. You teach to the level instead of over people's heads. You can see progress as people move up. And you know exactly where your champions are, because they are the ones approaching the top.
A workforce does not become capable in a workshop. It climbs, one stage at a time, and the climb is only visible if you bother to measure it.
Sources: FSCA & Prudential Authority, Artificial Intelligence in the South African Financial Sector (Nov 2025); MIT NANDA, State of AI in Business (2025).
