Hong Kong’s AI education blueprint is a start – but schools need more
Hong Kong’s AI education blueprint is a start – but schools need more
Hong Kong’s schools stand at a critical juncture as artificial intelligence (AI) reshapes learning. The government’s Blueprint for Digital Education Development in Primary and Secondary Schools, launched last month just days ahead of Digital Education Week, sets out a clear vision with students as the focus, teachers as professionals, schools as the base and society as a partner.
It has incorporated concerns and suggestions raised by educators and researchers, including an AI pedagogical framework, a plan to build a shared resource platform, and progressive AI literacy training for teachers.
Yet recent evidence from schools shows there is still some gap before achieving meaningful AI integration in education.
Our Hong Kong Foundation’s survey, conducted between July and December last year, found that while two-thirds of teachers say they integrate AI tools in their classrooms, the figure varies sharply by subject. About 89 per cent of information and communication technology teachers reported teaching their students to use AI tools, while 70 per cent of languages and science teachers did so. In comparison, only around 44 per cent of mathematics teachers reported integrating AI in their teaching, with the percentage falling to 40 per cent for visual arts, music and history.
Digital education brings the opportunity to rethink how every subject is taught and how learning is assessed. Publishing a blueprint is only the first step. The blueprint should trigger subject-wide curriculum renewal, with AI literacy built into every subject’s learning objectives, with teachers supported to use AI in areas where it deepens thinking, improves practice or offers new perspectives.
That is a demanding task for already-busy teachers. Simply urging more AI use risks either token gestures or quiet resistance.
Professional development must therefore focus on pedagogy and providing concrete, subject-based examples: a language teacher seeing how AI can support reading and writing without compromising the ability to learn grammar; a mathematics teacher seeing how AI can help students analyse logical errors without providing an exact answer; a history teacher seeing how AI can bring competing narratives into the same discussion.
The Education Bureau should go beyond broad guidelines and create a structured library of case studies on effective AI-related lesson plans across subjects. This practical resource will enable teachers to adapt proven models and foster consistent implementation across schools.
Building such a resource requires genuine partnership. Universities can contribute research on pedagogy and help identify what works. Technology firms can offer usable tools and real-world perspectives. Schools can provide the testing ground, showing what survives contact with everyday constraints. With this kind of collaboration, AI can move from one-off showcases to sustained, evidence-based practice.
If teaching methods change, assessment must also change. In a world where AI can generate essays and imitate creative work, the traditional “results-only” assessment no longer captures what students can truly do. Clinging to old models risks rewarding those most adept at outsourcing tasks, not those with a strong understanding.
Schools will need to place more weight on learning processes. That means examining how students frame and refine prompts, how they critique machine-generated answers and how they detect and correct errors. Oral exams, structured discussions and live presentations can capture these capabilities. These rely on real-time reasoning and interaction, which are far harder to delegate to software.
Over time, AI can also help build fairer, more continuous assessments by tracking students’ learning journeys, recording how they approach tasks and respond to feedback. With clear guard rails and professional judgment, this can reduce the pressure on single high-stakes examinations.
For students, this creates the space to rediscover the satisfaction of exploring ideas and solving problems, rather than living in the shadow of a single score. It widens the conversation about what educational success should mean in the long term.
All this leads back to a fundamental question: what is the purpose of education in an intelligent era? The answer is not to compete with machines, but to return to the human core of education – sparking curiosity, cultivating critical thinking and safeguarding humanistic values. As AI takes over more routine work, it becomes even more important to nurture students who can ask good questions, work with others and make principled choices.
For the blueprint to succeed, authorities, school leaders, teachers, parents and the community must collaborate closely to review and refine education practice. This collective effort will help ensure technology supports educational goals, enabling students to thrive in the AI era.
Hong Kong’s schools stand at a critical juncture as artificial intelligence (AI) reshapes learning. The government’s Blueprint for Digital Education Development in Primary and Secondary Schools, launched last month just days ahead of Digital Education Week, sets out a clear vision with students as the focus, teachers as professionals, schools as the base and society as a partner.
It has incorporated concerns and suggestions raised by educators and researchers, including an AI pedagogical framework, a plan to build a shared resource platform, and progressive AI literacy training for teachers.
Yet recent evidence from schools shows there is still some gap before achieving meaningful AI integration in education.
Our Hong Kong Foundation’s survey, conducted between July and December last year, found that while two-thirds of teachers say they integrate AI tools in their classrooms, the figure varies sharply by subject. About 89 per cent of information and communication technology teachers reported teaching their students to use AI tools, while 70 per cent of languages and science teachers did so. In comparison, only around 44 per cent of mathematics teachers reported integrating AI in their teaching, with the percentage falling to 40 per cent for visual arts, music and history.
Digital education brings the opportunity to rethink how every subject is taught and how learning is assessed. Publishing a blueprint is only the first step. The blueprint should trigger subject-wide curriculum renewal, with AI literacy built into every subject’s learning objectives, with teachers supported to use AI in areas where it deepens thinking, improves practice or offers new perspectives.
That is a demanding task for already-busy teachers. Simply urging more AI use risks either token gestures or quiet resistance.
Professional development must therefore focus on pedagogy and providing concrete, subject-based examples: a language teacher seeing how AI can support reading and writing without compromising the ability to learn grammar; a mathematics teacher seeing how AI can help students analyse logical errors without providing an exact answer; a history teacher seeing how AI can bring competing narratives into the same discussion.
The Education Bureau should go beyond broad guidelines and create a structured library of case studies on effective AI-related lesson plans across subjects. This practical resource will enable teachers to adapt proven models and foster consistent implementation across schools.
Building such a resource requires genuine partnership. Universities can contribute research on pedagogy and help identify what works. Technology firms can offer usable tools and real-world perspectives. Schools can provide the testing ground, showing what survives contact with everyday constraints. With this kind of collaboration, AI can move from one-off showcases to sustained, evidence-based practice.
If teaching methods change, assessment must also change. In a world where AI can generate essays and imitate creative work, the traditional “results-only” assessment no longer captures what students can truly do. Clinging to old models risks rewarding those most adept at outsourcing tasks, not those with a strong understanding.
Schools will need to place more weight on learning processes. That means examining how students frame and refine prompts, how they critique machine-generated answers and how they detect and correct errors. Oral exams, structured discussions and live presentations can capture these capabilities. These rely on real-time reasoning and interaction, which are far harder to delegate to software.
Over time, AI can also help build fairer, more continuous assessments by tracking students’ learning journeys, recording how they approach tasks and respond to feedback. With clear guard rails and professional judgment, this can reduce the pressure on single high-stakes examinations.
For students, this creates the space to rediscover the satisfaction of exploring ideas and solving problems, rather than living in the shadow of a single score. It widens the conversation about what educational success should mean in the long term.
All this leads back to a fundamental question: what is the purpose of education in an intelligent era? The answer is not to compete with machines, but to return to the human core of education – sparking curiosity, cultivating critical thinking and safeguarding humanistic values. As AI takes over more routine work, it becomes even more important to nurture students who can ask good questions, work with others and make principled choices.
For the blueprint to succeed, authorities, school leaders, teachers, parents and the community must collaborate closely to review and refine education practice. This collective effort will help ensure technology supports educational goals, enabling students to thrive in the AI era.
