The AI Education Race Has Started — and South Africa Has Not Left the Building

Singapore is integrating AI tools at primary level with teacher supervision. China has mandated it nationally. The UAE has deployed 1,000 trained teachers to deliver it from kindergarten. South Africa's national AI policy does not exist. The race for the workforce of 2030 is already underway.

Sign up for my Substack daily AI newsletter here.

See my AI Training course portfolio for corporate Business Leaders here.

Article link: https://open.substack.com/pub/johanosteyn/p/the-ai-education-race-has-started?r=73gqa&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true

Follow me on LinkedIn: https://www.linkedin.com/in/johanosteyn/

At Saint Anthony’s Canossian Secondary School in Singapore, a fifteen-year-old student named Anaiya Singhvi uses Google’s NotebookLM to explain chemistry concepts she finds difficult — the invisible architecture of molecules and atoms that is hard to visualise in a textbook. The school’s chemistry department uses AI to help students identify which topics need more focus. “I’ve been using AI to help me with that,” she told ABC News Australia in a feature. She is not exceptional in this. She is representative of what a generation of students in a country that has made deliberate, systematic choices about AI education now looks like.

The question that Anaiya’s experience raises for South Africa is not whether AI will be part of her future workforce. It already is part of her education. The question is what her South African equivalent’s education looks like — and what it will look like in five years, when both of them are competing in the same global labour market.

CONTEXT AND BACKGROUND

The regional picture that surrounds Singapore’s classroom experience is the most important context for South African education policymakers, school leaders, and board members with any relationship to skills development and human capital strategy.

China has mandated AI integration across every stage of its national K-12 education system — from primary school through to the final year of secondary education. The integration covers AI fundamentals, data and algorithms, applications, and ethical considerations. It reaches every child in the world’s most populous country. The UAE introduced artificial intelligence as a mandatory subject across all public schools from kindergarten through Grade 12 in the 2025-2026 academic year, deploying approximately 1,000 specially trained teachers to deliver a structured curriculum. For the 2026-2027 academic year, the UAE Ministry of Education formalised the subject as permanent infrastructure rather than a pilot. India is rolling out AI and Computational Thinking from Grade 3 onwards across all schools beginning in the 2026-2027 academic year, aligned with its National Education Policy 2020. Singapore offers structured, teacher-supervised AI tools for primary school students through its Student Learning Space platform — a deliberate infrastructure choice that builds AI feedback loops into a governed, localised portal rather than allowing unregulated access to external models.

Each of these decisions was made within the last two years. Each reflects a national conclusion that AI literacy is not an elective capability for the workforce of the future but a baseline requirement — and that the time to build it is now, in classrooms, at scale, with properly trained teachers and a governed curriculum.

South Africa’s equivalent decision has not been made. South Africa’s Draft National AI Policy, approved by Cabinet in March 2026 and gazetted for public comment in April, was withdrawn weeks later after at least six of its 67 academic citations were found to be AI-generated fabrications — journals that did not exist, articles that had never been written, authors who had never produced the research attributed to them. The country that has the continent’s most developed technology sector, the most established university research base, and the deepest experience with AI deployment across financial services, mining, and healthcare does not currently have a functioning AI governance framework. Its education system does not have an AI education policy. And the students in its classrooms are being prepared for a labour market that their Asian peers are already operating in.

INSIGHT AND ANALYSIS

The most instructive story in the Asian AI education landscape is not the success. It is the setback. South Korea allocated the equivalent of approximately 850 million US dollars for a rapid national AI digital textbook rollout — an initiative that was forced into a drastic scale-back and multi-year suspension just months into its deployment. The technology was not the problem. The system outpaced teacher readiness and triggered severe institutional pushback from parent groups and teacher unions over screen time and systemic unpreparedness. South Korea’s setback is the most valuable data point available to South African education policymakers right now: it is a precise demonstration of the sequencing error that produces expensive failure — and it happened in one of the highest-performing education systems in the world.

That last point carries particular weight for South Africa. South Korea’s rollout failed on top of a foundation of high educational performance — a system where basic numeracy, literacy, and computational thinking rates already rank among the world’s highest on international benchmarks including TIMSS and PIRLS. The system was strong enough to experience a sequencing failure and survive it as a setback rather than a collapse. Singapore and the UAE avoided the same failure by investing in teacher preparation before scaling student access. South Korea outpaced its teachers. Singapore trained its teachers first.

South Africa’s foundational context is fundamentally different. A large majority of South African Grade 4 learners cannot read for meaning in any language — a foundational literacy crisis that has been documented, lamented, and inadequately addressed for more than a decade. Introducing a rapid, scaled AI curriculum on top of that foundational deficit would not produce a South Korean-style setback. It would exponentially widen the domestic digital divide, adding an AI literacy gap to a foundational literacy gap in the same students, in the same classrooms, at the same moment. The sequencing question for South Africa is therefore more complex than the Asian experience describes — it is not only about training teachers before deploying technology. It is about the order of foundational investments required before AI education can deliver on its promise rather than compound existing inequality.

South Africa’s dominant institutional response to AI in schools has been risk management — plagiarism detection, academic integrity policies, tool bans. The Asian frame is capability building — how do we ensure our students can use these tools critically, ethically, and productively. Those are not competing approaches to the same problem. They are responses to fundamentally different questions about what education is for in an AI era. And the countries answering the second question are producing the graduates that the labour market of 2030 will reward.

UNESCO’s analysis of generative AI in Asia-Pacific education adds the cultural and linguistic dimension that most international comparisons understate. Most generative AI models are trained primarily on Western data, creating contextual and cultural relevance gaps for Asian students — gaps that are more acute for African students learning in African languages with African cultural contexts.

The Asian experience is instructive but not directly transferable. South Africa’s AI education challenge is compound: the digital access gap, the foundational literacy gap, the teacher training gap, the language and cultural relevance gap, and the policy vacuum all require sequenced, coordinated responses. The Asian experience shows what happens when any one of those gaps is not addressed before the technology is deployed at scale. In South Africa, all of them exist simultaneously.

IMPLICATIONS

For South African boards, executives, and institutions with any relationship to education, skills development, or human capital strategy, the Asian AI education picture carries three specific implications.

The first is about the workforce pipeline. The students who will enter the South African labour market in 2030 and 2031 are currently in school. The AI literacy — or the absence of it — they are building now will determine their productivity, their adaptability, and their competitiveness in a labour market where AI is already a standard tool in every sector that employs graduates. The organisations that are planning their human capital strategy for the next decade without accounting for the AI literacy level of the graduates they will be hiring are planning with an incomplete picture of what those graduates will and will not know how to do. Given South Africa’s foundational literacy challenges, that picture should also account for the possibility that the graduates most exposed to AI tools in school will be those in well-resourced, urban institutions — compounding the existing inequality in graduate capability that most South African employers already manage.

The second is about the private sector’s role. South Africa’s public education system will not solve the AI literacy gap without government policy. But it will also not solve it with government policy alone. The private sector organisations most dependent on AI-capable graduates — technology companies, financial services, professional services, mining — have a direct interest in the AI literacy level of the pipeline that feeds their entry-level roles. The most practically useful contribution those organisations can make to the policy conversation is not advocacy for AI education in principle but specific, funded engagement with the teacher development, foundational literacy, and curriculum design questions that determine whether an AI education policy produces outcomes or documents.

The third is about urgency and sequencing. South Korea’s experience demonstrated that scale without preparation produces expensive failure even in high-performing systems. South Africa’s foundational literacy crisis means the sequencing challenge is more acute — the preparation required before effective AI education is possible includes investments in reading for meaning that the country has not yet made at scale. The organisations, institutions, and policymakers that use the current policy vacuum not as a reason to wait but as an opportunity to develop the preparation infrastructure that a future policy will require will be better positioned than those that wait for the policy before beginning the work.

CLOSING TAKEAWAY

Anaiya Singhvi is fifteen. She uses AI to understand chemistry. Her school uses AI to identify which topics she needs to focus on. In a decade, she will be a working professional in an economy that has been systematically building AI literacy since before she reached secondary school. The equivalent South African fifteen-year-old is in a school without an AI education policy, in a country without a national AI governance framework, being taught that AI is primarily a risk to academic integrity — and in many cases, in a school where the foundational literacy required to benefit from any technology-enhanced education has not yet been secured.

The AI education race has started. The countries that will produce the most capable AI-literate workforces of 2030 are the ones making decisions now — about foundational prerequisites, about teacher training, about curriculum design, and about the sequencing of access and governance. South Africa has not yet made those decisions. The race does not pause while the starting position is debated.

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