top of page

South Africa is debating AI in education while ignoring the real emergency

The real crisis is foundational learning and dropout, and the AI era will magnify the consequences of every unresolved gap.




There is a growing fascination with “AI in schools” conversations: personalised tutors, automated marking, smarter lesson planning, and dashboards that promise insight at scale. I understand the appeal. But I want to argue something simpler and more uncomfortable: if we are serious about our children’s future, we must stop treating AI as the main storyline and start treating basic literacy, numeracy, and retention as the emergency. 


Muhammad Coovadia’s searing Daily Maverick piece, SA’s “education” system is warehousing children, not educating them, captures this moral reality with rare clarity: the applause for pass rates can become complicity when the system is quietly failing millions.


CONTEXT AND BACKGROUND

Coovadia’s argument is not abstract. He describes older learners who cannot write a basic sentence, the normalisation of catastrophic averages, and the way “success” is framed around a single percentage while the missing learners are treated as statistical noise. That concern is echoed by civil society. The Zero Dropout Campaign has repeatedly pushed the country to look at throughput, not just matric results, because the pipeline tells the truth the podium often avoids.


This is not only a moral issue, but it is also a measurement problem. The Department of Basic Education’s own reporting and planning documents reveal a system heavily shaped by targets, monitoring, and institutional incentives. Meanwhile, independent research and journalism keep returning to the same foundational gap: early literacy is in crisis, and children who don’t learn to read for meaning early rarely recover later.


INSIGHT AND ANALYSIS

Now add AI to this picture. In a healthy system, AI can help teachers prepare materials, translate content, summarise reading, support learners who need extra practice, and reduce administrative burden. In a broken system, AI can become a cosmetic layer over structural failure. It can generate the appearance of progress while leaving the underlying capability gap untouched.


This is where the AI era changes the stakes. Synthetic content is exploding, persuasion is becoming cheaper, and misinformation is more scalable. If large numbers of learners cannot read for meaning, they are not only disadvantaged in the job market, but they are also more vulnerable to manipulation. The most urgent “AI skill” for our children is not prompt-writing. It is discernment: reading, reasoning, checking, and resisting emotional bait. That’s why the “warehouse” critique is not separate from AI; it is the foundation of our democratic resilience.


International assessments and reporting underline the inequality dimension. Le Monde’s coverage of South Africa’s performance in international testing points to deep, persistent disparities across the system. Afrobarometer’s work adds another layer: public trust in government performance on education is mixed, and that trust gap matters when you’re asking society to buy into new technologies in classrooms.


IMPLICATIONS

So what should leaders do with this? First, we should be honest about sequencing. AI in education should be a second-order conversation, not the headline, until we can show consistent progress on foundational learning and learner retention. Daily Maverick reporting on targeted interventions to reduce dropout aligns with this: if learners are falling out of the system, no classroom technology strategy will reach them.


Second, we need accountability that cannot be spun. Publish pass rate alongside throughput, year after year, and treat dropout as a national crisis with the same seriousness we reserve for other systemic shocks. Media Monitoring Africa’s Back-to-school review reminds us how often the lived reality of schools is shaped by conditions and access problems that make learning harder before a lesson even begins.


Third, any AI procurement or pilot must be tied to child-centred outcomes, not vendor promises. The Star’s reporting on the dropout crisis around the matric class of 2025 is a warning: if the pipeline is leaking, technology can’t be the victory lap.


CLOSING TAKEAWAY

I am not against AI in education. I am against distraction dressed up as innovation. Coovadia is right to challenge the applause when the ship is taking on water, because the children who cannot read for meaning, who are socially promoted without support, or who disappear before matric, are not a footnote.


In the AI era, weak foundations do not merely limit opportunity; they magnify vulnerability. If we want our children to thrive in a world of synthetic content and automated persuasion, the rescue mission starts with basics: literacy, numeracy, retention, and dignity. AI can help, but only after we stop pretending the emergency is somewhere else.


Author Bio: Johan Steyn is a prominent AI thought leader, speaker, and author with a deep understanding of artificial intelligence’s impact on business and society. He is passionate about ethical AI development and its role in shaping a better future. Find out more about Johan’s work at https://www.aiforbusiness.net


 
 
 

Comments


Leveraging AI in Human Resources ​for Organisational Success
CTU Training Solutions webinar

bottom of page