What SEA Educators Actually Think About AI in the Classroom (And What the Data Gets Wrong)

Marielyn Wong
3 days ago
7 min read

There is a story that gets told constantly in EdTech circles: teachers are afraid of AI. They see it as a threat to their jobs, a shortcut that undermines learning, or simply another technology initiative that will fade before they finish the training. The narrative is tidy, and it lets institutions off the hook — if the problem is teacher resistance, then the solution is awareness campaigns and mindset workshops.

The data from Southeast Asia tells a different story entirely.

The "Resistance" Narrative Is a Red Herring

According to EdTech Hub's March 2026 findings on AI in education across Southeast Asia, approximately 86% of teachers are already using AI tools — not because they were told to, but because they found them useful [1]. ChatGPT and similar tools are being used to plan lessons, generate classroom materials, and reduce administrative load. Students in the region are accessing AI on their smartphones. Governments from Manila to Kuala Lumpur are investing in digital infrastructure.

This is not a portrait of a profession in denial. It is a profession that adopted a technology without waiting for permission.

And yet the same research highlights that much of this engagement is happening "with limited training, guidance, or institutional frameworks to support it." Teachers are experimenting in the dark. They are running fast without a road. The problem is not that they are unwilling — it is that no one built the conditions for them to succeed.

What the Data Actually Shows

Across the region, the gap between enthusiasm and effective implementation is stark and consistent.

In the Philippines, the Department of Education launched AGAP.AI in January 2026 — the country's first nationwide program for responsible AI integration in basic education, targeting 1.5 million students, teachers, and parents [2]. Microsoft's partnership under the ARAL program is rolling out AI tools to 3,000 teachers across 1,500 schools throughout 2026 [3]. These are real commitments. But they are also responses to a pre-existing reality: teachers in Philippine public schools were already figuring out AI on their own, often without formal guidance on ethics, accuracy, or pedagogy.

Malaysia moved faster at the policy level. Its National Education Plan 2026–2035, launched in January 2026, embeds AI and digital literacy across the full system, from primary school to university. All 20 public universities introduced Google Gemini for Education tools this year, reaching nearly 600,000 students [4]. Yet a Tatler Asia analysis noted that the teacher training dimension of Malaysia's AI push remains uneven — well-resourced urban schools are moving ahead, while rural and lower-income schools face infrastructure gaps and fewer trained facilitators [5].

Singapore is arguably the most structured in the region. The MOE's AI-in-Education (AIEd) Framework, aligned with the EdTech Masterplan 2030, sets clear pedagogical principles for AI use and ties it to the National AI Strategy [6]. The National Institute of Education's AI@NIE programme and a new Certificate in AI for Education (launched July 2025) are building teacher capacity with applied, judgment-based training. Singapore demonstrates what it looks like when a system invests in professional development seriously — but it is also a city-state with unique resource concentration that cannot simply be copy-pasted to a province in Mindanao or a secondary school in rural Sarawak.

The Digital Education Council's Global AI Faculty Survey (2025) found that while 63% of K-12 teachers globally have incorporated generative AI into their teaching — up 12% year on year — the quality and confidence of that use varies enormously depending on the support structures behind them [7]. In the Asia-Pacific context, the region's 35.3% CAGR for AI in education is impressive on paper [8]. But compound annual growth rates measure investment, not impact. They do not capture whether a teacher in Cebu feels equipped to explain to a student why an AI-generated answer might be wrong.

The Real Gap Is Not Willingness. It Is Design.

EdTech Hub's 2026 research is clear on this point: educators and policymakers across Southeast Asia expressed "uncertainty about emerging AI trends and a need for reliable, contextualised guidance" [1]. The gap is not motivational. It is structural.

Three specific gaps appear repeatedly across the region:

1. Time to experiment. Teachers who are already managing overcrowded curricula, large class sizes, and administrative demands do not have discretionary time to explore tools, iterate on approaches, or reflect on what worked. AI experimentation requires slack in the system. Most SEA classrooms do not have it.

2. Culturally and linguistically relevant tools. The dominant AI tools are designed primarily in English, for Western pedagogical contexts. A Filipino teacher working in Filipino and regional languages, or a Malaysian teacher navigating a multilingual classroom, is not simply using a tool — they are adapting something that was never designed for their reality. EdTech Hub's topic briefs on marginalised learners in the region flag this as a systemic concern: without inclusion by design, AI risks deepening existing inequities rather than closing them [9].

3. Structured, subject-specific professional development. This is the most consistently underinvested area across the region. Most AI training for teachers currently takes the form of generic workshops — a half-day session on what AI is, a demo of a few tools, and a certificate. This is not professional development. It is orientation.

What Effective AI Professional Development Actually Looks Like

The distinction matters, and it is worth being direct: a workshop that explains how ChatGPT works is not the same as professional development that builds teacher judgment about when and how to use AI in a specific subject, with specific learners, in a specific cultural context.

Effective AI PD for teachers looks like this:

It is subject-specific. A science teacher needs different scaffolding than an English teacher or a school counsellor. Generic training flattens those differences.
It is ongoing, not episodic. Competence with AI tools builds through iteration — trying something, observing results, refining. A one-time workshop cannot achieve this.
It includes failure cases. Teachers need to see where AI gets things wrong, particularly in local language and cultural contexts. Confidence in AI use is not the same as uncritical acceptance.
It connects to curriculum, not just tools. The most effective models anchor AI use to actual lesson planning and assessment design, not abstract demonstrations.
It builds teacher agency, not dependency. The goal is for teachers to make informed choices about when AI adds value and when it does not — not to become power users of a particular platform.

Singapore's NIE programme models several of these principles. So does the structured mentoring component of DepEd's AGAP.AI rollout. These are the approaches worth studying and scaling — not replicating the tool access without the pedagogical framework around it.

What Institutions and Policymakers Should Do Differently

If the problem is implementation design rather than teacher readiness, the solutions are correspondingly specific.

Stop framing AI adoption as a change management problem. Teachers are already adopting AI. The question is whether they are doing it well. Framing this as a resistance issue misdirects attention and resources.

Fund time, not just tools. Professional learning requires protected time. Any AI initiative that does not build in structured, recurring time for teachers to learn, practice, and debrief is underfunded by design.

Demand localisation from EdTech vendors. Procuring AI tools that are not tested in local languages, curricula, or cultural contexts is not cost-effective — it shifts the adaptation burden onto individual teachers. Policymakers and procurement bodies should require evidence of local relevance before deployment at scale.

Build regional knowledge-sharing infrastructure. EdTech Hub's work through the ASEAN–UK SAGE Programme is a step in this direction — creating mechanisms for countries in the region to learn from each other rather than importing frameworks wholesale from the US or UK [1]. This should be expanded and sustained.

Measure teacher capability, not just access. The dominant metrics for AI in education count devices, subscriptions, and training hours. Almost none of them assess whether a teacher can use AI to make a lesson better. That measurement gap enables a great deal of theatre.

What This Looks Like in Practice

The recommendations above describe what the region needs from its institutions and policymakers. The same standard applies to EdTech vendors — and it is worth naming what it actually looks like when a platform is built for this context rather than retrofitted onto it.

Noodle Factory is a Singapore-based EdTech platform built specifically for educators in Southeast Asia. It supports multilingual content creation — relevant to teachers working in Filipino, Bahasa Malaysia, Bahasa Indonesia, and multilingual classroom environments — and is designed around the curricula, regulatory norms, and pedagogical realities of the region. It was not adapted from a Western product. It was built here, for here.

What distinguishes the model is less the technology than the partnership approach. Rather than selling a licence and leaving schools to figure out implementation, Noodle Factory works directly with institutions to co-build AI into their existing workflows — supporting teachers through the transition rather than handing them a tool and walking away. For a region where the primary barrier is not teacher willingness but institutional support and contextualised guidance, that distinction matters.

It is not the only model worth watching, but it is an example of the kind of vendor that policymakers and procurement bodies should be looking for when they demand localisation: one that can demonstrate, specifically, how its product performs in local languages, against local curricula, and with the structural support to help teachers use it well.

The Closing Argument

The teachers of Southeast Asia are not waiting to be convinced about AI. They are waiting for their institutions to catch up with them — to provide the time, the tools, the context-specific guidance, and the pedagogical frameworks that would let their existing curiosity translate into consistent, high-quality classroom practice.

Getting this wrong is not a neutral outcome. A region that invests heavily in AI access but lightly in teacher support will produce a generation of learners who interact with AI tools they neither understand nor can critically evaluate. That is a literacy crisis deferred, not solved.

The question is not whether SEA educators are ready for AI. It is whether the systems around them are ready to take that readiness seriously.

Sources

EdTech Hub — AI in Education Across Southeast Asia: What's Working on the Ground (March 2026): https://edtechhub.org/2026/03/30/ai-in-education-across-southeast-asia-whats-working-on-the-ground/
DepEd AGAP.AI launch and Microsoft ARAL programme overview: https://news.microsoft.com/source/asia/2026/02/03/deped-and-microsoft-accelerate-learning-recovery-and-ai-literacy-for-filipinos/
Philippines AI integration in public schools — teacher readiness and best practices: https://www.academia.edu/130240135/THE_INTEGRATION_OF_ARTIFICIAL_INTELLIGENCE_AI_IN_TEACHING_SCIENCE_AMONG_PUBLIC_SCHOOLS_IN_THE_PHILIPPINES_BEST_PRACTICES_CHALLENGES_AND_READINESS_OF_TEACHERS
Malaysia AI in education 2026 — leading the shift: https://www.tatlerasia.com/power-purpose/others/ai-in-malaysian-education-2026
UNICEF Digital Education — Malaysia AI and student retention: https://www.unicef.org/digitaleducation/stories/future-ready-and-resilient-how-ai-keeping-malaysian-students-school
Singapore MOE AI-in-Education Framework and EdTech Masterplan 2030: https://www.moe.gov.sg/education-in-sg/educational-technology-journey/edtech-masterplan/artificial-intelligence-in-education
Digital Education Council — Global AI Faculty Survey 2025: https://www.digitaleducationcouncil.com/post/digital-education-council-global-ai-faculty-survey
AI in education global statistics and Asia-Pacific growth data: https://www.demandsage.com/ai-in-education-statistics/
EdTech Hub Topic Brief — AI in Southeast Asia: Marginalised Learners: https://docs.edtechhub.org/lib/ZAIZ22IV/download/2666S75U