In 2029, PISA will formally assess Media and Artificial Intelligence Literacy for the first time. They are calling it MAIL. Every participating country's 15-year-olds will be tested on their ability to engage with, manage, and think critically about AI systems.
That is three years away. The students who will sit that assessment are in your school right now.
The OECD and European Commission have already published a draft framework for how to teach this. It is called the AILit Framework, built in partnership with Code.org, and it covers three scenarios: students using AI independently, students and teachers using it together during lessons, and teachers using it for planning and assessment.
The framework is clear about one thing. AI literacy is not a Computing subject. It cuts across everything. Humanities. Sciences. Languages. Art. PE. The whole curriculum.
What the framework actually says
The OECD framework is built around the idea that AI literacy should be interdisciplinary. Not bolted onto existing ICT lessons. Not delivered in a standalone "AI week." Woven into the subjects students are already studying.
That sounds obvious when you say it. It is very hard to do in practice. Because it means every teacher needs to understand how AI connects to their subject, and most teachers have never been shown that connection.
A Science teacher who teaches the scientific method is one step away from teaching students how to evaluate AI-generated claims. A Geography teacher covering data visualisation is sitting on a lesson about how AI systems interpret and present information. An English teacher asking students to analyse persuasive writing could be asking them to analyse AI-generated persuasive writing.
The connections exist. Teachers just need someone to draw the line.
The Council of Europe went further
While the OECD focused on pedagogy, the Council of Europe approached AI literacy through human rights. At its plenary session in Strasbourg in March, the Steering Committee for Education approved a recommendation that puts the "human dimension" first. Ethics. Democratic participation. Critical thinking. Cultural awareness. Environmental impact.
Their argument is straightforward. If students learn to use AI without learning to question it, you have not taught them literacy. You have taught them compliance. Real AI literacy means understanding what the technology does to power structures, to truth, to autonomy, and to the environment.
That recommendation is expected to be formally adopted in June. It will push member states to integrate a three-dimensional model into national curricula: the human dimension, the technological dimension, and the practical dimension.
The human dimension comes first. Not the technical one. That ordering is deliberate.
What this means for your school
Your school will be compared internationally on AI literacy in 2029. The framework that defines what "good" looks like has already been drafted. The competencies are public. The assessment design is underway.
If your school's AI literacy strategy is currently "we'll get to it," you are already behind. Not behind the school down the road. Behind entire countries.
India has mandated AI modules from Year 6. South Korea is restructuring university courses because students relied on AI too much. The EU has made AI literacy a legal obligation for employers. Malaysia has trained 80,000 learners through a national programme.
The question is not whether PISA 2029 is coming. It is whether your students will be ready when it arrives.
And that does not start with a whole-school training day in 2028. It starts with one teacher, one lesson, one connection between their subject and AI literacy. Repeated across the school, subject by subject, term by term.
That is exactly the problem AILitKit was built to solve. You do not need to understand AI to use it. You need to understand your subject. AILitKit draws the line between the two.
Matthew Wemyss is the founder of AILitKit and IN&ED, and author of AI in Education: An Educator's Handbook.