Beijing has done what most education ministries have spent three years debating: it has made artificial intelligence literacy compulsory, cradle to career, backed by a national timetable and tied explicitly to industrial policy. China’s Ministry of Education has unveiled an action plan to build a comprehensive AI literacy system spanning every stage of schooling and lifelong learning, with a 2030 completion target. The announcement lands amid a wider scramble among the world’s largest school systems — from Abu Dhabi to New Delhi to Singapore — to decide, in real time, what an AI-literate citizen should look like by the time they leave school. Malaysia, despite outsized ambitions in the same industry, is notably absent from that list.
The gap is no longer theoretical. It is now measurable in national gazettes, teacher-certification exams and university admissions policy, and it is emerging as one of the defining education stories of 2026 across nine of the world’s most consequential school systems.
Context: from pilot programmes to national mandates
For most of the past decade, AI in education meant scattered pilot projects — a chatbot trial here, a coding elective there. That era ended in 2026. China’s plan folds AI into teacher qualification exams and certification requirements, with the stated aim of a fully built literacy infrastructure by 2030, and officials have been explicit that the education reform doubles as industrial policy: the country’s 15th Five-Year Plan (2026–2030) calls for securing a leading position in AI industry applications, and a school system that has never touched the technology cannot supply that workforce.
The UAE moved first among early adopters, introducing AI as a mandatory subject across all public schools from kindergarten through Grade 12 starting in the 2025–2026 academic year, delivered by roughly 1,000 specially trained teachers and formalised for 2026–2027 under the title “Artificial Intelligence and Technology.” India is following at even greater scale: AI and Computational Thinking will enter all schools from Grade 3 onward beginning in the 2026–27 academic year, aligned with the National Education Policy 2020. Singapore has taken a narrower, deeper approach — integrating AI modules into primary-level computer science and committing to AI training for teachers at every level, including trainee teachers, by 2026.
South Korea offers the cautionary counter-example. Seoul poured the equivalent of US$850 million into an AI textbook initiative that collapsed within four months of launch, after the rollout outpaced teacher readiness and infrastructure capacity. The lesson being drawn across ministries this year is that national ambition without implementation scaffolding is not merely wasteful — it can set a reform back years.
Malaysia’s paradox: chip exporter without a classroom mandate
Malaysia exports around 13% of the world’s semiconductors and hosts nearly half of all planned data-centre capacity across Southeast Asia. Its prime minister has committed RM2 billion to a sovereign AI cloud. Yet there is still no national mandate for AI in school curricula and no announced timeline for AI to enter teacher certification, unlike the requirement China introduced this year.
That is not to say nothing is happening. Malaysia’s National Education Plan 2026–2035, launched in January, commits to embedding AI, STEM and digital literacy from primary school through university, and in 2026 all 20 public universities rolled out Google’s Gemini for Education tools across academic platforms, reaching nearly 600,000 students and tens of thousands of lecturers. Technical and vocational institutions are issuing micro-credentials in automation, robotics and data analytics, and the Thirteenth Malaysia Plan (2026–2030) pushes closer industry-education collaboration. What is missing is the structural, primary-school-upward mandate that China, the UAE and India have already legislated — and analysts argue that gap matters precisely because Malaysia’s semiconductor and data-centre ambitions require a domestically trained AI-literate workforce within the decade, not after it.
Singapore and Indonesia: two speeds of the same transition
Singapore’s posture has been described by education officials as “precision over breadth” — narrower in scope than China’s system-wide mandate, but layered and sequenced. In May 2026, Education Minister Desmond Lee announced that autonomous universities, polytechnics and the Institute of Technical Education will embed baseline AI competencies through compulsory modules for all incoming students by 2027, building on a pathway that begins in secondary school and junior college. Separate competency frameworks exist for universities versus polytechnics and ITE, reflecting different vocational endpoints. The initiative sits under Singapore’s National AI Council, formed to help the city-state treat AI as a strategic economic asset rather than a classroom novelty.
Indonesia has opted for a gentler on-ramp. From the 2025–2026 academic year, Indonesian primary and secondary schools began rolling out AI and coding as elective subjects, with fifth-grade primary and junior-high students spending roughly two hours a week on the subject, rising to five hours at senior secondary level, while the Ministry of Primary and Secondary Education finalises the implementing regulation. Officials frame the goal explicitly in workforce terms: producing students ready to become, in the words of one ministry director, “Indonesia’s future technology leaders” — language that echoes China’s industrial framing almost verbatim, despite the two countries choosing very different degrees of compulsion.
Why this now counts as a global governance story, not just an edtech one
Four threads make 2026 the inflection point rather than merely another year of incremental edtech adoption.
First, the mandates are now backed by law and budget, not white papers. China’s plan folds AI into teacher licensing; the UAE has hired and trained a dedicated teaching cadre; Singapore has set a hard 2027 deadline for higher-education-wide competency modules. These are enforceable commitments with consequences for non-compliance, not aspirational frameworks.
Second, the divergence is becoming a competitiveness signal that markets and multilateral bodies are starting to track. A country’s AI-in-schools policy is increasingly read as a proxy for its medium-term position in the global AI value chain — which is precisely why Malaysia’s absence from the mandate list, despite its chip-manufacturing weight, has drawn outside scrutiny.
Third, the failure case is now public and specific. South Korea’s collapsed textbook programme gives every finance ministry a concrete cautionary tale about sequencing capital spend against teacher readiness — a lesson likely to shape how Pakistan, Indonesia and other budget-constrained systems phase their own rollouts.
Fourth, and most consequential for households: the compulsory AI subject a student encounters in Grade 3 in India, or in kindergarten in the UAE, or not at all in a Malaysian public school, will shape university admissions, vocational tracking and early labour-market sorting well before 2030. Parents comparing school systems across ASEAN, South Asia and the Gulf are, for the first time, comparing AI curricula as a distinguishing feature rather than a bonus.
The unresolved question — one none of these nine markets has fully answered — is whether compulsory AI literacy actually produces better economic outcomes, or whether it simply produces better-branded school systems. South Korea’s collapse suggests speed alone guarantees neither.
