A calculator does not weaken a mathematician. It weakens a pupil who never learned to multiply. That distinction decides whether artificial intelligence strengthens a school or hollows it out. The technology has already entered the staff room, the office of the principal and the administration block.

By Silas Mwaudasheni Nande

The arithmetic of an ordinary school week

Every teacher knows the shape of the week. There is the teaching, and then there is everything that surrounds the teaching. There are schemes of work and lesson preparations. There are worksheets, marking, mark sheets, remedial plans and reports. The second category consumes the evenings and the weekends.

The pressure is not imagined, and it is not peculiar to Namibia. UNESCO has recorded that sub Saharan Africa must recruit about 15 million teachers by 2030. The same assessment recorded one trained teacher for every 58 learners at primary level across the region. At secondary level the figure was about 43 learners for every trained teacher.

A school does not solve that ratio by exhortation. Teachers in this country already work hard, and most work beyond the hours they are paid for. The number of teachers is fixed by the establishment. The number of learners is fixed by the community. Only one variable remains open to a school, and that is the time consumed by tasks that never required professional judgement in the first place. The scarce resource in a Namibian school is not goodwill. It is the time of a qualified teacher.

What the evidence actually shows

Opinion on this subject is plentiful. Measurement is rarer, and it deserves more weight. The most substantial measurement available was published by Gallup, in partnership with the Walton Family Foundation, on 24 June 2025.

The survey covered 2,232 public school teachers. It was conducted between 18 March and 11 April 2025, through a nationally representative panel, with a margin of error of 2.5 percentage points. The findings were specific.

• Sixty per cent of teachers reported using artificial intelligence tools in their work.
• Thirty two per cent reported using such tools at least once a week.
• Teachers who used the tools weekly reported saving 5.9 hours in a working week.
• Across a school year of 37.4 weeks, that saving amounts to about six weeks of time returned.

The saving was not spread evenly across the duties of a teacher. Respondents reported time saved most often on the making of worksheets, at 84 per cent. Assessments followed at 83 per cent, administrative work at 81 per cent, lesson preparation at 80 per cent, and marking and feedback at 79 per cent.

Reported quality also rose. Majorities said the quality of their work improved, ranging from 57 per cent for marking and feedback to 74 per cent for administrative work. Teachers who used the tools weekly were roughly twice as likely as occasional users to report much higher quality.

Two cautions belong beside those figures. These are perceptions reported by teachers, and not outcomes measured independently. The survey was conducted in the United States, and no equivalent study has been published for Namibia. The figures indicate a direction. They do not constitute a promise.

Evidence from the African continent

One rigorous study has been conducted closer to home. The World Bank reported on a pilot in Edo State, Nigeria, in a paper published on 9 January 2025. Learners attended an after school programme for six weeks, between June and July 2024, supported by a generative artificial intelligence tutor.

The measured gain was about 0.3 of a standard deviation. The authors described that gain as equivalent to nearly two years of ordinary learning, achieved in six weeks. Compared against a database of education interventions tested by randomised controlled trials in developing countries, the programme outperformed 80 per cent of them.

One point of clarity is needed here. The six weeks in the Nigerian study refers to the length of the programme itself. It has no connection to the six weeks of saved working time reported in the Gallup survey above. The first concerns how long learners were taught. The second concerns how much time teachers recovered across a year.

That result is striking, and it should still be treated with care. It concerns one subject area, one cohort and a short period. It has not been replicated in Namibian conditions. It is nevertheless the strongest African evidence currently available, and it points the same way as the survey data.

What this means for the teacher

Planning and daily preparation

The first honest use of artificial intelligence is drafting. A teacher can produce a first version of a scheme of work against the syllabus in minutes rather than hours. The draft will be imperfect. It will nevertheless be a starting point, and starting is the part that consumes the evening.

The same applies to daily preparation. A teacher may ask for three different introductions to a lesson on photosynthesis. A teacher may ask for a worksheet at three levels of difficulty, for the fast learners, the middle group and those who struggle. A teacher may ask for a reading passage to be rewritten at a simpler reading level, without losing the content.

One caution is specific to Namibia. The national readiness assessment recorded a shortage of training data in indigenous languages. Output in Oshiwambo, Otjiherero, Khoekhoegowab or Rukwangali should therefore be checked closely by a first language speaker before it reaches a learner.

Assessment of learners

Assessment is where the time cost is heaviest and where the gain is largest. A teacher can draft a question paper aligned to stated objectives, build a marking memorandum, and produce a table of specification showing the spread of cognitive levels. A teacher can convert a column of raw marks into an item analysis that shows which question defeated the class.

A boundary must be drawn here, and it is not negotiable. The teacher awards the mark. The machine may draft the question, may suggest a rubric and may summarise a pattern. Professional judgement on the performance of a named child remains with the professional who knows that child.

Teaching itself

In the classroom the value is explanation. A teacher who has explained long division three times, and has still not reached one child, can ask for four further approaches within a minute. A teacher can request analogies drawn from cattle, from mahangu fields or from the school tuck shop, rather than from a foreign textbook. Relevance is a pedagogical asset, and it has always been expensive to produce.

What this means for the principal

Evidence suggests that school leaders have adopted these tools faster than classroom teachers. A study by the RAND Corporation, published on 11 February 2025, examined the 2023 and 2024 school year. It found that 25 per cent of teachers had used artificial intelligence for instructional planning or teaching. It found that nearly 60 per cent of principals had used such tools in their work.

The reason is not difficult to identify. The working day of a principal is dominated by documents. There are annual plans, budgets, circulars, minutes, reports to the circuit, reports to the region, speeches and correspondence with parents. Every one of those begins with a blank page.

Planning and strategic management

A principal can convert a list of priorities into a structured annual plan, with objectives, indicators, responsible officers and target dates. A principal can ask for the weaknesses in a draft strategic plan to be identified before the management team meets. A principal can prepare an agenda, and then convert rough notes into formal minutes on the same afternoon.

Artificial Intelligence in Namibian Schools,

Monitoring and evaluation of programmes

This is the most underused application in school leadership. A principal can submit three years of subject results and ask for the trend, the outliers and the subjects in decline. A principal can ask which grade lost the most ground between the first and the second reporting period. A principal can request a class visit schedule that covers every teacher once a term without clashing with examinations.

A principal can also convert a folder of lesson observation notes into a summary of recurring strengths and recurring weaknesses. That summary then becomes the agenda for staff development, grounded in evidence rather than impression.

A warning belongs here. A principal who tables an unchecked report will eventually table an error, and the error will be public. Every figure must be traced back to the source document before the report is signed.

What this means for the administrative officer

The administrative officer is the post most often forgotten in this discussion, and the post where relief would be felt fastest. The duties are repetitive by design, and repetition is precisely what these tools handle well.

• Drafting routine correspondence, acknowledgements, invitations and standard letters to parents.
• Converting handwritten notes into formal minutes and circulars.
• Writing spreadsheet formulas for registers, stock records and petty cash reconciliation.
• Summarising long circulars from the circuit or the region into the three points that require action.
• Preparing inventory descriptions, filing indexes and standard forms.

One prohibition governs all of it. Personal information about a learner or a member of staff must never be entered into a public tool. That includes identity numbers, home addresses, medical details, disciplinary records and examination material.

This matters more in Namibia than in many countries. The national readiness assessment, launched on 14 August 2025, recorded that the Data Protection Bill had not been finalised. Where the law does not yet protect a learner, the professional standards of the school must.

The position of Namibia has already been taken

Some staff rooms still debate whether artificial intelligence belongs in a school. That debate has been overtaken by policy.

Hon. Dr Sanet Steenkamp, the Minister of Education, Innovation, Youth, Sport, Arts and Culture, set out the position in The Namibian on 2 July 2026. The Minister stated plainly that artificial intelligence cannot replace teachers. The approach adopted is human centric, with technology serving educational goals while human interaction and professional judgement are preserved.

The Minister was equally clear about the practical value. Artificial intelligence can assist with lesson planning, with assessments and with administrative work. Automating routine tasks frees teachers for mentoring and for the emotional development of learners. The Minister also required that consequential decisions affecting learners must not rest on a machine alone.

The structural position matches. On 20 August 2026 the Ministry launched the 2025 and 2026 Norms and Standards Framework for Schools. The framework prioritises digital literacy and the integration of information and communication technology, including applications of artificial intelligence. Deputy Minister Hon Dino Ballotti reminded the launch that standards are ultimately about people.

Implementation has also begun. A pilot at Luderitz Secondary School, launched in July 2025, ran as a five month research project with connectivity support from Telecom Namibia. The stated intention is expansion to 25 schools and 2,500 learners. The question before Namibian schools is no longer permission. It is competence.

Internationally, guidance already exists. UNESCO published its Guidance for Generative AI in Education and Research on 7 September 2023. It calls for a human centred approach, for the protection of data privacy, and for an age limit on independent conversations between children and these platforms. It warns that most countries lack national regulation, leaving institutions unprepared to validate the tools they adopt.

The honest objection: does it make us lazy?

This objection deserves a serious answer rather than a dismissal, because there is evidence behind it.

Michael Gerlich published a study in the journal Societies in 2025. It surveyed 666 participants in the United Kingdom and included 50 interviews. The study found a significant negative correlation between frequent use of artificial intelligence tools and critical thinking. The relationship was substantially mediated by cognitive offloading, meaning the habit of handing mental work to an external system.

Two findings from that study matter especially for schools. Younger participants showed greater dependence on the tools and lower critical thinking scores. Higher educational attainment was protective, and was associated with better critical thinking regardless of how often the tools were used.

A second paper, published in Frontiers in Psychology in 2025, calls this the cognitive paradox of artificial intelligence in education. Its authors conclude that artificial intelligence must be designed as an enabler rather than a substitute. That paper reports research from the University of Pennsylvania. Learners using a chatbot answered 48 per cent more practice problems correctly. The same learners scored 17 per cent lower on conceptual understanding.

Three qualifications are owed to the reader. Correlation is not causation. The Gerlich study surveyed the general population of one country, not teachers, and not Namibians. The Pennsylvania finding is reported here as it appears in the Frontiers paper, because the original study was not consulted directly.

Even with those qualifications, the danger is real, and it is specific. It is not a danger of using the tools. It is a danger of a particular way of using them. A teacher who copies a lesson plan unread has abandoned the profession, with or without a machine.

Where the machine does the typing, the formatting, the first draft and the repetitive production, no thinking is lost. The thinking was never located there. Where the machine does the analysis, and the person accepts the output without examining it, something real is surrendered.

The professional therefore keeps the judgement and delegates the labour. A teacher who asks for a draft worksheet, and then adapts it for a particular class, has lost nothing and gained an hour. The reverse is also true, and it does not require a machine.

It is also worth remembering that the same anxiety accompanied earlier tools. The duplicating machine, the photocopier, the pocket calculator and the desktop computer each arrived with predictions of decline. None of them produced lazy teachers. Each of them returned time, and the profession absorbed them and moved on.

Ten rules for a school that intends to use these tools well

Enthusiasm is not a policy. The following rules are offered as a starting point for any school that decides to proceed.

1. The professional signs the work. Responsibility for any document rests with the person whose name appears on it.
2. No personal information about a learner or a member of staff is ever entered into a public tool.
3. Every fact, date, figure, name and reference is verified before use. These systems produce confident errors.
4. The machine drafts. The human decides. That order is never reversed.
5. No mark is awarded, and no placement decision is taken, by a machine alone.
6. Use is declared wherever honesty requires it, including in work submitted for assessment.
7. Staff are trained before they are expected to comply. Budget for that training.
8. The school adopts a written policy of one page, and reviews it every year.
9. Learners are taught to use these tools openly and critically, rather than punished for discovering them alone.
10. The school reviews the effect after one term, and records what worked and what failed.

The tenth rule carries particular weight in Namibia. The RAND study of 2025 found that 45 per cent of principals reported having a school or district policy on artificial intelligence. It also found that more than 80 per cent of learners said their teachers had never explicitly taught them how to use these tools for schoolwork. Adoption ran ahead of guidance, and Namibian schools have the opportunity not to repeat that error.

A breakthrough, properly understood

There is a further reason for Namibian schools to move deliberately rather than slowly. The RAND research found that teachers and principals in higher poverty schools reported markedly lower adoption than those in wealthier schools. Principals in the poorest schools were half as likely to provide guidance as principals in the wealthiest.

That pattern is a warning for a country with the rural and urban disparities that Namibia carries. A tool that saves six weeks of professional time each year will widen the gap between schools if only some schools learn to use it. The remedy is training and access, delivered first where the need is greatest.

Artificial intelligence will not replace the teacher who stands before 40 learners in Ondobe. It cannot notice that a child has not eaten. It cannot telephone a guardian, sit with a grieving pupil, or decide that a boy needs encouragement rather than punishment. Those acts are the profession, and they are not automatable.

What the technology can do is take back the Sunday afternoon spent formatting a worksheet. It can take back the evening spent typing minutes. It can take back the hours spent building a question paper from nothing. On the available evidence that amounts to roughly six weeks of a working life each year.

Six weeks is not a trivial gift. Returned to a teacher, it becomes preparation, remediation and rest. Returned to a principal, it becomes time in classrooms rather than behind a desk. Returned to an administrative officer, it becomes accuracy instead of backlog. That is the case for embracing this technology, and it rests on evidence rather than on enthusiasm.

Artificial intelligence is here. It will not wait for the comfortable, and it will not punish the careful. Namibian educators should approach it as they would approach any powerful instrument entering a school. Learn it, govern it, set its limits in writing, and then put it to work.

Silas Mwaudasheni Nande is a Namibian school principal and an analytical writer on governance and public policy. He is a doctoral candidate in education, researching leadership of information and communication technology in rural schools. The views expressed are personal.

The post Delegate the Labour, Keep the Judgement: Artificial Intelligence in Namibian Schools appeared first on The Mt Kenya Times.