The EDB AI Literacy Framework Explained: What It Means for Your Child's Education

When the Education Bureau publishes policy documents, they tend to be written for educators and administrators rather than parents. The EDB's work on AI literacy in schools is no exception. The language is measured, the frameworks are structured, and the direct implications for your child's actual school experience are buried several paragraphs deep.

I've spent time working through the key documents, and I want to translate them into something more useful for families trying to understand what Hong Kong is — and isn't — doing to prepare students for an AI-shaped future.

What the EDB has actually published

The EDB's approach to AI literacy has evolved across several documents and initiatives rather than arriving as a single comprehensive policy. The key strands are worth understanding separately.

The first strand is AI as a subject of study — understanding what AI is, how it works in broad terms, its applications and limitations. This sits within the ICT and Computing curricula at secondary level, and has been gradually strengthened.

The second strand is AI as a tool for learning — using AI-assisted applications to support education across subjects. This is where the guidance is less settled. The EDB has issued general guidance on responsible use of generative AI, but implementation varies significantly across schools.

The third strand — and the most important for long-term outcomes — is AI literacy as a cross-curricular capability. The idea that understanding AI isn't just for ICT students but is relevant to every subject and every student. This is the most ambitious strand and also the one where the gap between stated intention and classroom reality is widest.

The competency areas in plain language

The framework identifies several areas of AI literacy that students should develop. Translated into parent-friendly terms, these are:

Understanding how AI systems make decisions. Not programming an AI from scratch, but grasping that when an algorithm recommends a video, approves a loan, or assesses a job application, there are rules and patterns behind that decision — and those rules reflect the choices (and sometimes the biases) of the people who built the system.

Recognising AI-generated content. As deepfakes, AI-written text, and synthetic images become more prevalent, students need the ability to think critically about whether content is what it claims to be. This is digital literacy for an era that is categorically different from the one most parents grew up in.

Applying AI tools responsibly. Using AI assistance in ways that are honest, that respect others' work, and that don't create harmful outputs. This is partly about academic integrity and partly about broader ethical behaviour with powerful tools.

Understanding data and privacy. AI systems learn from data. The data they learn from often belongs to people. Understanding how personal data flows through the systems students interact with daily — social media, apps, devices — is foundational to informed participation in digital life.

What this looks like (and doesn't look like) in schools

Here's the honest part. The framework is sensible. The implementation is patchy.

Schools with strong ICT departments and teachers who've invested in understanding these areas are moving quickly and doing genuinely interesting work. I've seen computing classes where S3 students are building simple machine learning classifiers, understanding what training data is and why biased data produces biased results. This is excellent.

Schools where computing is treated as a secondary priority — and there are many — are largely doing the minimum. Students complete their ICT assignments, learn the vocabulary, and move on. They might be able to define "machine learning" on an exam without having any genuine understanding of what it means.

The variance is wide enough that I wouldn't assume your child is receiving meaningful AI literacy education just because it's on the EDB framework. The question to ask is: "What does your child actually do in ICT/computing lessons? What have they built or explored?"

What you can do at home

You don't need to be a technologist to support AI literacy. Some of the most effective things parents can do involve conversation rather than instruction.

When your child encounters an AI-generated output — a recommendation on YouTube, a suggestion from a search engine, a chatbot — ask: "How do you think that system decided to show you this?" The goal is to replace the sense that digital systems are neutral and magical with the understanding that they are built, they have purposes, and they sometimes make mistakes.

When your child reads news or information online, occasionally ask: "Is there any chance this wasn't written by a human?" Start pointing to the signals — suspiciously uniform sentence structures, no byline, content that is technically accurate but somehow doesn't feel observed from experience. This kind of critical reading is learnable.

Follow the EDB's general guidance on screen time and technology use at home, but don't confuse screen time limits with AI literacy education. Reducing time with devices doesn't automatically produce children who understand the devices they're using. The understanding comes from guided engagement, not avoidance.

The bigger picture

Hong Kong is competing with cities and countries that are investing heavily in technical education. Singapore, Shenzhen, Taipei — our neighbours are taking digital and AI literacy seriously at a systemic level. The EDB's framework is a genuine attempt to keep pace.

Whether it translates into actual capability for the students sitting in classrooms today depends on implementation — on school leadership, on teacher training, on whether families treat these skills as important enough to reinforce at home.

The framework is the floor. What your child actually develops depends on what happens above it.