Webinar Tackles AI's role in Shaping the Education Landscape Across Southeast Asia

Ghada Qureshi, Jazzlyne Gunawan, Sangay Thinley | EdTech Hub

Southeast Asia. Ed Tech Hub (30 March 2026) - Across Southeast Asia, AI is arriving in classrooms faster than education systems can responsibly respond. Students are already using AI tools on their smartphones, teachers are experimenting with new applications, governments are investing in digital infrastructure, and technology providers are expanding rapidly into the region. Yet much of this engagement is happening with limited training, guidance, or institutional frameworks to support it. The question for education systems is no longer whether AI will be adopted, but how its use can be guided so that it is safe, evidence-informed, and beneficial for those most at risk of being left behind. 

On 10 March 2026, EdTech Hub hosted a webinar on How AI Is Shaping the Education Landscape Across Southeast Asia: What’s Working on the Ground, with support from the ASEAN–UK SAGE Programme. The session launched EdTech Hub’s series of AI in Education Topic Briefs, exploring five key dimensions of AI adoption across the region: strategic partnerships, marginalised learners, ethical governance, girls’ education, and the evolving role of teachers. Developed in response to knowledge gaps identified through stakeholder consultations across the region, the briefs translate emerging global evidence into practical guidance for Southeast Asian education systems. 

This session brought together researchers, practitioners and policymakers to discuss key insights and pressing concerns about data protection, the invisibility of learners with disabilities, and the risks of scaling AI without the appropriate governance frameworks or cross-sector collaboration.

WATCH THE WEBINAR HERE

Speakers

• Natalja Rodionova, Sisters of Code
• Dr. Sahawarat Polahan, Equitable Education Fund
• Bela Nurul Fadhilah, ASEAN Foundation
• Dr. Nurhasmiza Sazalli, Universiti Teknologi Malaysia
• Dr. Kruakae Pothong, London School of Economics.

Key insights from the discussion

1. Across Southeast Asia, teachers are already using AI, often without formal training or institutional support.
2. Expanding access to technology is necessary but not sufficient; invisible barriers persist, and without deliberate design, AI systems risk reinforcing existing learning inequalities, especially for those most marginalised.
3. Girls are increasingly engaging with AI tools, but they benefit most when learning takes place in structured environments that support digital literacy and online safety.
4. Data protection gaps and weak governance frameworks risk leaving children exposed to privacy violations, commercial exploitation, and other harms. 
5. Without regional coordination and inclusion by design, AI risks fragmenting policy and deepening existing inequities.

Key Takeaways

These key takeaways highlight practical lessons on using AI in education, focusing on teachers, equity, safety, and systems-level collaboration.

Build teachers’ confidence and capacity to use AI effectively

Across Southeast Asia, many teachers are already experimenting with AI tools in their daily work. Findings from EdTech Hub’s Role of Teachers brief suggest that around 86% of teachers are already using AI tools — a point echoed during the discussion. Tools such as ChatGPT are being used to help plan lessons, generate classroom materials, and reduce administrative workload. Much of this experimentation, however, is taking place without formal training or institutional guidance. Teachers in lower-resourced settings are often navigating this shift largely on their own, covering hidden costs such as subscriptions, devices, or connectivity.

Dr. Nurhasmiza Sazalli from Universiti Teknologi Malaysia addressed concerns that teachers may be reluctant or sceptical about using AI in their practice. In her experience, the main barrier is often confidence rather than the technology itself. Teachers often question whether AI-generated outputs are reliable enough for classroom use or whether they are prompting the tools in ways that produce meaningful results. Her approach is to prioritise hands-on training where teachers generate materials and experiment directly. Within her programmes, educators have used tools like NotebookLM to design posters, create infographics, and draft lesson materials. 

As she observed, "once they see how quickly AI can support their teaching tasks, their confidence increases significantly.”

These experiences suggest that teachers are already shaping how AI enters classrooms. The challenge for education systems is to ensure that training and institutional support evolve alongside this rapid experimentation.

Design AI that removes structural barriers for marginalised learners

The Marginalised Learners brief highlights that expanding access to technology is often only the first step in integrating AI into education. As the discussion emphasised, access alone rarely translates into meaningful learning outcomes.

During the webinar, Dr. Sahawarat Polahan, Research Director at Thailand’s Equitable Education Fund (EEF), illustrated this through a national initiative that distributed free high-speed internet SIM cards to more than 400,000 eligible students. Of those, 113,000 registered, yet only 18,000 actively used them for educational purposes. As he explained, “even when we remove the financial barriers to internet access, other invisible barriers remain.” These include limited digital literacy, a lack of relevant learning content, and weak motivation to use technology for learning.

Rather than beginning with technology, EEF first seeks to understand the environments in which learners live and study. The organisation is developing AI systems drawing on student risk profiles, community-level data collected through Thailand’s Sirindhorn Anthropology Centre, which documents local histories and socio-economic conditions, and scholarship opportunity data to enable personalised pathway recommendations for disadvantaged students. This aligns with EdTech Hub’s education- and evidence-led philosophy: beginning with the lived realities of learners, grounding decisions in contextual data, and using evidence to guide how technology is designed and deployed to support continued learning.

The discussion reinforced a broader lesson reflected throughout the brief: the effectiveness of AI in education does not depend solely on technological capability, but more so on how well systems reflect the realities of the communities they are intended to serve.

Support girls’ engagement with AI through literacy and safeguards

The Girls’ Education and Empowerment brief highlights a dual reality. AI is opening new opportunities for girls to engage with digital learning and future employment pathways, yet confidence gaps and social norms continue to shape participation in technology across Southeast Asia. In many contexts, STEM and digital fields are still widely perceived as male-dominated domains, discouraging girls from engaging with advanced technologies. Practitioners are also beginning to observe how AI tools, particularly conversational systems such as large language models, may be shifting this dynamic by lowering barriers to entry. Natural language interfaces make complex technologies feel more approachable, allowing girls to engage without needing advanced technical skills at the outset. 

Natalja Rodionova, founder of Sisters of Code in Cambodia, explained, “AI, especially large language models, is opening the door for many girls who previously felt technology was too complicated. When they can use natural language, they become curious and much more willing to start learning.”

AI is also expanding entry points into technology. Rodionova noted that in some Sisters of Code programmes 90% of students did not own computers, yet 70% were already using ChatGPT through their smartphones. However, exposure and experimentation alone are not enough to produce meaningful impact without a structured educational approach. Rodionova stressed that students need guidance on how AI systems work, how to understand the data they rely on, and how to apply critical thinking when using them. Ensuring that curiosity translates into meaningful learning requires structured AI literacy and opportunities to engage critically with the technology.

Isla Gilmore from the UK Mission to ASEAN also reminded participants that “girls and women are disproportionately victims of online abuse and safeguarding concerns.” As AI becomes more embedded in digital learning environments, ensuring girls can engage safely online is essential to prevent new technologies from reinforcing existing risks. At the same time, evidence from programmes such as ASEAN-UK SAGE suggests that girls are highly receptive to digital safety training, making AI literacy framed through a safety lens a particularly effective entry point for reaching young women.

The challenge now is ensuring that access evolves into agency, through the literacy, safeguards, and learning environments that allow girls to engage with AI not just as users, but as contributors to the digital futures taking shape around them.

Embed governance, privacy, and child protection safeguards into AI systems

As AI tools become embedded in learning environments, questions of trust, accountability, and child protection become unavoidable. Findings from the Ethical Governance of AI in Education brief highlight uneven compliance with national data protection standards across EdTech platforms. Of the eight companies reviewed, only three specified age thresholds for defining a child and only two had age verification practices in place. Security and accountability measures were often weak, and very few provided AI-specific disclosures.

Dr. Kruakae Pothong, Research Fellow at the Digital Futures for Children Centre at the London School of Economics, placed these findings in a broader global context. Drawing on Human Rights Watch research, which found that many EdTech platforms exposed children to privacy risks or engaged in practices that potentially violated their rights. Without stronger governance and clearer guidance for developers, she warned, “children become sitting ducks for commercial exploitation.”

Participants pointed to practical resources that can help translate child protection principles into design practice, including the UNICEF RITEC toolbox, the Digital Futures for Children Playful by Design framework, and the Child Rights by Design toolkit. Indonesia’s adoption of an age-appropriate design code was also highlighted as an example of stronger regulatory protection. 

The discussion also underscored that governance is not simply a regulatory exercise but a condition for trust. As speakers emphasised during the session, scale without trust will not succeed. Gaps in transparency, data protection, and child safeguarding can quickly erode public confidence, regardless of how advanced the technology itself may be. As AI becomes embedded in everyday learning tools, ultimately, governance will determine whether these technologies deepen trust in education systems or quietly erode it.

Build partnerships for responsible AI adoption across sectors and borders

The integration of AI in education is fundamentally a systems challenge. The Strategic Partnerships brief notes that AI initiatives across Southeast Asia typically rely on collaborations between governments, technology providers, researchers, and civil society organisations, each contributing different expertise, resources, and pathways to scale. Bela Nurul Fadhilah of the ASEAN Foundation noted that when these actors work towards shared goals, scale and impact become possible; when they do not, even well-resourced initiatives stall.

Indonesia’s rollout of coding and AI embedded as curriculum priorities reflects the scale of investment some governments are making. Whether it translates into equitable outcomes will depend on the broader ecosystem surrounding it. The brief also highlights that many education-focused AI initiatives remain short-term and project-based, raising questions about how partnerships can evolve beyond pilots to support long-term impact. As Dr. Majah-Leah Ravago of SEAMEO INNOTECH cautioned, without a shared education-specific guidance, the region risks fragmented policymaking, uneven standards, inconsistent safeguards, and widening gaps between systems.

Dr. Ravago also highlighted ongoing efforts to develop a Southeast Asian regional policy framework on AI in education; a co-creation process that the topic briefs are designed to inform. Whether the region moves from shared evidence to shared action will depend on the partnerships, governance, and political will that education systems are able to build together.

Highlights from the Chatbox:

Throughout the webinar, participants also contributed actively in the chat – engaging with the panelists who shared further resources to the participants, and raising additional reflections on digital literacy, AI design, and the governance challenges emerging as AI tools enter classrooms.

• The digital divide is increasingly about literacy, not just access. Learners who benefit most from generative AI are those who can question outputs and work around accuracy limitations; a skill that itself requires structured support.
• AI literacy and media literacy need to go hand in hand. Participants emphasised the need for skills to identify misinformation and verify AI-generated content, noting that neither can be treated as an add-on to existing curricula.
• Consent in school-based EdTech environments is rarely straightforward. When digital tools are required for learning, students may have little real choice about using them, raising questions about what meaningful consent looks like in practice.
• Child-centred design must be built in from the start, not added later. Participants highlighted age appropriateness, alignment with pedagogy and developmental needs, and thoughtful classroom integration as non-negotiables for any AI tool entering an education setting.

Together, the discussion underscored a central lesson: while AI technologies are advancing rapidly, their impact in education will ultimately depend on how thoughtfully they are integrated into real learning environments. Supporting teachers, addressing structural barriers, safeguarding learners, and strengthening regional collaboration will be essential if AI is to expand opportunity rather than deepen existing inequalities.

This article is for educational purposes only. All credit and rights of this report remain with the original copyright holders such as owners, authors, and publishers of the material. More information on this can be found on our Additional Disclaimers page.