Scaling AI Education Tools for Young Learners in Hong Kong

Susanna is Associate Vice President (Quality Assurance), Professor in the Department of Psychology, and Executive Co-director of the Academy for Educational Development and Innovation at The Education University of Hong Kong. She previously served as Associate Dean (Quality Assurance and Enhancement) and Associate Head (Programmes) within the university.

She holds a BSocSc and an MPhil in Psychology, as well as a doctorate in educational psychology, all from The University of Hong Kong. Her research spans language and reading development, reading intervention, and technology-supported learning. She has commercialised a parent-child learning kit for second-language acquisition and secured funding to develop a social robot for early language learning.

In an interview with M R Yuvatha, Editor, Asia Education Review, Susanna argues that research on how children learn particularly in language and reading development, must now embrace educational technology and AI as essential tools, while ensuring that the human relationships at the heart of early education remain protected and strengthened.

Artificial intelligence tools are becoming embedded in classrooms across Hong Kong and beyond, but a growing body of research is challenging a common assumption in edtech: that more AI automatically produces better learning outcomes. Research in early childhood and language education points to a different conclusion. The real opportunity lies not in how advanced the technology is, but in how carefully it is designed, supervised, and integrated into a child's learning journey particularly during the earliest and most formative years of education.

AI as a Tool, Not a Substitute for Guidance

AI is fundamentally a tool one that requires human involvement to function well in education. This holds true across the education spectrum, but the level of involvement required changes with age. For K-12 students, responsibility for guiding AI use falls squarely on teachers and school leadership, who must build not just AI literacy in students, but critical thinking and creativity alongside it.

University students tend to explore AI tools more independently, largely because they have grown up as digital natives. This shifts the burden onto institutions and educators, who must be well prepared and equipped with AI tools and knowledge, while also designing tasks and assessments capable of measuring genuine learning outcomes rather than AI-assisted output.

AI is not positioned as a replacement for teachers, parents, or human judgment, but as a tool whose value depends entirely on design quality and the strength of human guidance surrounding it.

The Equity Question Extends Beyond the Classroom

Scaling AI-powered learning tools across schools and communities’ raises equity concerns the sector cannot ignore. In Hong Kong specifically, a well-resourced educational environment across both public and private sectors has helped limit equity gaps at the school level, supported by consistent government investment in teacher development across K-12 and university contexts.

The deeper equity issue sits outside the education system entirely it rests with parents and communities. People across different workplaces and backgrounds need access to AI education as well, positioning equitable AI access as a societal responsibility that industry, government, and families must address together, not an issue schools can solve alone.

Younger Children Need Carefully Designed Platforms, Not Open Access

The clearest and most urgent finding concerns the earliest years of childhood: direct, unsupervised interaction between young children and large language models should not happen. Policy in many education systems already reflects this position, because young children typically cannot distinguish or critically evaluate AI-generated output on their own.

For children in early childhood and lower primary years, AI platforms require careful, purpose-built design rather than open-ended chat access to a language model. Older students can gradually take on more independent AI use, but that independence must be paired with continued educational guidance specifically, clarity on which tasks are appropriate to delegate to AI and which require independent effort. Every AI-based learning activity demands deliberate design, a standard with direct implications for how edtech companies build products for this age group.

Two outcomes anchor confidence that AI-powered learning tools are making a genuine difference, personalization and efficiency. Personalization has long been an education goal that individual teachers struggle to deliver at scale, given constraints on time and capacity. Properly designed AI systems can track student progress and deliver appropriate feedback and scaffolding at a scale no individual teacher can match.

This defines AI's core value proposition in education, in the same amount of time, students can learn more. That is the value of AI.

A clear line separates AI as a simple question and answer tool from AI as a genuine educational instrument. Effective AI systems pair with data analytics rather than functioning merely as a question-answering interface, the narrower use case is too simplistic for education's actual demands a critique with direct implications for how the edtech industry designs and markets its products.

The Future of Learning Lies in AI-Powered Personalization

Personalization stands out as the most significant shift AI will bring to early childhood and primary education across Hong Kong and the wider region in the coming years. Tasks and feedback tailored to individual student progress, delivered without the time constraints limiting individual teachers, mark a meaningful departure from one-size-fits-all instruction.

AI also stands to reshape after-school learning substantially. Static worksheets requiring extended waits for teacher feedback can give way to engaging, game-like exercises with immediate responses, compressing a feedback loop that has long slowed learning reinforcement outside the classroom.

Most significantly, AI offers a potential answer to a persistent inequity in education, the gap in home support between children of more educated parents and those without that advantage. Parents with stronger educational backgrounds have historically provided more coaching and support at home. Appropriately designed, accessible AI tools could narrow that gap substantially, extending meaningful learning support to families regardless of parental education level an outcome with significant implications for how the sector prioritizes accessibility and affordability in product design.

This body of research rejects both uncritical AI enthusiasm and reflexive resistance, establishing a calibrated middle standard for the industry to build around. AI is not positioned as a replacement for teachers, parents, or human judgment, but as a tool whose value depends entirely on design quality and the strength of human guidance surrounding it.

For younger children, that means tightly controlled, purpose-built platforms rather than open language-model access. For older students, it means structured independence paired with continued educational guidance. For the equity question, it means recognizing that meaningful AI access cannot be solved by schools alone, but requires support extending into homes and communities well beyond the classroom.

Also Read: Asia's Digital Literacy Push: What's Working, What's Not

Forging Ahead!

As education systems across Hong Kong and the wider region continue integrating AI into early learning, this research sets a clear, evidence-grounded standard for the sector, AI's promise will only be realized if human oversight, thoughtful platform design, and educational intent remain central to how it is built and deployed.