The short answer
Schools without computer labs can teach AI literacy effectively. A 2025 meta-analysis of 37 studies found unplugged activities match computer-based ones at the primary level. The barrier is not hardware. It is that teacher training assumes a computer-first classroom most schools in Southeast Asia and South Asia do not have. The question is how to prepare teachers to deliver AI literacy with what they already have.
India's CBSE board now mandates AI and Computational Thinking from Class 3 for its 28,000+ affiliated schools, with state boards expected to follow. The Philippines is rolling out Project AGAP.AI across basic education. The UAE has deployed AI fundamentals to every public school from Kindergarten through Grade 12.
Here is what those mandates collide with on the ground.
In India, 35.3% of schools have no functional computers. 36.5% have no internet connection. Rural schools trail urban ones by 29 percentage points in connectivity (UDISE+ 2024-25). In the Philippines, only 22% of primary students sit in a classroom where their teacher has access to a computer, even though 75% of schools technically own some (SEA-PLM, 2019). The learner-to-computer ratio in Philippine primary schools was 412 to 1 the last time it was formally measured (UNESCO, 2012, via PIDS 2024). The metric has not been updated since, because authorities shifted focus to per-lab provisioning.
Those are not statistics about schools that need a budget increase. They describe a structural reality where the standard model of AI education, open a laptop and use a tool, does not apply. And yet, these are exactly the schools where teachers are now being told to teach AI.
Does AI literacy require a computer lab?
The evidence says no, at least not for the age groups where mandates are hitting hardest.
A 2025 meta-analysis published in Educational Technology Research and Development examined 37 studies comparing unplugged activities (no computers) to plugged activities (with computers) for developing computational thinking in students. At the primary school level, the effects were statistically comparable. In game-based learning contexts, unplugged activities outperformed plugged ones. And for interventions lasting more than four weeks, the gap disappeared entirely.
Unplugged curricula already exist that put this into practice. Northwestern University's AI Unplugged project (NSF-funded, CC-BY-NC 4.0, free) offers 13 activities covering supervised learning, neural networks, AI ethics, and data privacy using only paper, dice, and card games. ISTE's unplugged AI activities offer similar starting points. A 2026 CHI paper documents a framework for "Teaching AI by Hand" designed specifically for underserved contexts.
The materials are real. The pedagogy is sound. What is missing is the teacher who knows how to use them.
Why has teacher training not caught up?
Three reasons, none of which are about budgets or bandwidth.
Teacher training assumes computers. The training materials that accompany AI literacy mandates are overwhelmingly designed for computer-equipped classrooms. The Philippines' Project AGAP.AI plans AI-powered platforms for teaching and assessment. India plans to deliver AI teacher training through NISHTHA modules and video-based resources. These are useful for the 64.7% of Indian schools that have functional computers. They do not reach the other 35.3%. And they do not equip teachers to teach when the lab is occupied, the power is out, or the internet is down, which in many schools is most of the time.
Free unplugged curricula exist but are invisible to teachers in SEA and South Asia. Northwestern's materials are in English, designed for Western classroom norms, and connected to no ministry syllabus in the Philippines, Malaysia, India, or Singapore. A Filipino teacher who searches for "AI activities without computer" in DepEd's resources will find nothing. The resources exist. The bridge between them and the teacher who needs them does not.
Localisation is the bottleneck nobody talks about. A 2026 study of WhatsApp-based teacher PD in Cameroon found that English-only content excluded Francophone colleagues from participation. The same dynamic applies across Southeast Asia: teachers in the Philippines teach in Filipino, Malaysian teachers in Bahasa Melayu, Indian teachers in Hindi and dozens of regional languages. AI literacy materials in these languages, unplugged or otherwise, barely exist. Content localisation, not device procurement, is the binding constraint on whether mandates reach the teachers who are supposed to deliver them.
How can teachers be trained without a computer lab?
The same infrastructure gap that blocks student-facing AI education also blocks teacher training. But the workarounds for teachers are, if anything, better evidenced.
Teachers already have the delivery platform in their pockets. In India, 59.2% of households own smartphones (53% rural, 70% urban). In the Philippines, penetration exceeds 70%. The Cameroon PD study tested structured training modules delivered via WhatsApp chatbot to 47 teachers. 77.1% already used WhatsApp daily. The chatbot was rated significantly more usable than traditional online forms (p=.046), and teachers described returning to the modules as classroom references between sessions. This is not a hypothetical channel. It is one that teachers have already adopted.
Unplugged activities are easier to learn to teach than software tools. A teacher training session on Northwestern's DataBites (students classify food items by hand to learn supervised learning) or EthiQuest (a card game on AI ethics) requires no IT setup, no troubleshooting, no login credentials. The teacher walks out ready to deliver the activity the next day. Compare that to training teachers on a platform that requires a lab booking, a stable connection, and 30 functioning machines.
The AI diary requires zero training and zero technology. Students log every interaction they have with AI-driven technology for one week, then discuss: what did the AI get right? What did it get wrong? Who decided how it works? Any teacher can facilitate this conversation. It covers UNESCO's "human-centred mindset" and "ethics of AI" dimensions. It works anywhere a student has seen a phone screen.
What should school leaders do when the mandate arrives before the devices?
Start with teacher readiness, not a materials shopping list. Before sourcing curricula or building channels, find out where your teachers actually stand on AI literacy. A school where teachers have never encountered AI concepts needs a different starting point than one where teachers use AI personally but have never brought it into a lesson. An infrastructure audit tells you what devices you have. A readiness assessment tells you what your teachers can do with them, or without them.
Treat unplugged activities as pedagogy, not a fallback. The evidence says these are not inferior at primary level. But the free curricula that exist are in English, designed for Western classrooms, and mapped to no Southeast Asian or South Asian syllabus. Schools that want to use them need localised, curriculum-aligned versions. Until ministries or PD providers bridge that gap, school leaders should be asking for it loudly.
Build your mobile-first PD channel before your computer lab. If your teachers have WhatsApp, you have a professional development delivery platform. The Cameroon study shows structured PD via chatbot outperforms traditional online forms in usability. Start with one AI literacy module per fortnight through the messaging app your teachers already check daily.
Tell your ministry what is missing. If your country's AI literacy mandate shipped without unplugged or mobile pathways, say so. Request localised materials in the language your teachers actually teach in. The mandates will not adapt to ground reality unless schools describe that reality clearly.
Frequently asked questions
Can teachers learn to teach AI without a computer lab?
Yes. Unplugged AI activities like card-based classifiers (DataBites) and ethical dilemma games (EthiQuest) require no IT setup and can be learned in a single training session. A 2026 study in Cameroon found that WhatsApp-based PD modules were rated significantly more usable than traditional online platforms (p=.046) by teachers who already used WhatsApp daily. The training channel does not need to match the classroom infrastructure the mandate assumes.
Does AI literacy require computers at the student level?
Not at primary level. A 2025 meta-analysis of 37 studies found unplugged activities produce comparable learning outcomes to computer-based ones for primary students (effect sizes of g = 0.501 vs g = 0.606). In game-based formats, unplugged activities outperformed computer-based approaches. The evidence supports hands-on, physical activities as a pedagogically effective method, not a compromise.
What is the student-to-computer ratio in Philippine and Indian schools?
In the Philippines, the ratio was 412 to 1 the last time it was formally measured (UNESCO, 2012), and only 22% of primary students have classroom access to a computer (SEA-PLM, 2019). In India, 35.3% of schools have no functional computers and 36.5% lack internet (UDISE+ 2024-25). Both countries have mandated AI literacy education.
Why are free AI curricula not reaching teachers in Southeast Asia?
The main barriers are language and curriculum alignment. Free resources like Northwestern's AI Unplugged are in English, designed for Western classroom norms, and connected to no ministry syllabus in the Philippines, Malaysia, India, or Singapore. Teachers in these markets teach in Filipino, Bahasa Melayu, Hindi, and dozens of regional languages. Content localisation, not device procurement, is the binding constraint.
What should schools without computer labs do about AI literacy mandates?
Start with a teacher readiness assessment to understand where your staff actually stands. Then build a mobile-first PD channel (WhatsApp-based modules work) and push your ministry for localised, curriculum-aligned unplugged materials. The AI diary exercise (students log and discuss daily AI interactions) requires zero technology and any teacher can facilitate it. The evidence shows unplugged approaches are pedagogically effective, especially at primary and lower-secondary levels. The gap is not in what works, but in who has been trained to deliver it.
If your school is navigating an AI mandate without a full computer lab, the question is not what devices to buy. It is what your teachers are ready to do now, and what structured pathway gets them to the next level. That is what thegurucool's TEACH-AI framework is built for: mapping where each teacher stands across 7 domains of AI readiness, and building from there, whether the school has a lab or not. Also see: AI Is Now in Asian Classrooms. Are Teachers Actually Ready for It?