If you only pay attention to one AI literacy framework, it should be this one. The OECD and European Commission published their AI Literacy Framework in May 2025 as a review draft, with the final version expected in 2026. It directly shapes what 15 year olds will be tested on in PISA 2029.
That connection to PISA is what makes AILit different from every other framework. UNESCO provides principles. The DOL provides practical categories. AILit provides the competences that will become an international assessment. When PISA results land in 2030, countries will be compared on their students' AI literacy. Schools that have started building these competences will show it. Schools that have not will show that too.
The four domains
AILit organises AI literacy into four domains. Each domain describes a different way students interact with AI.
Engage with AI is the largest domain. It covers recognising when AI is being used, understanding how AI systems work at a conceptual level, and evaluating what AI produces. This is where most schools will find their strongest existing connections. Any lesson where students evaluate information, question a source, or assess the reliability of a claim is already building Engage with AI competences. Eight competences sit within this domain.
Create with AI covers using AI tools to generate, refine, and iterate. This is not about using AI to do the work for you. It is about using AI as a creative partner while maintaining control of the quality and direction. Students learning to write clear prompts, refine outputs, and combine AI-generated material with their own ideas are developing Create with AI competences. Five competences sit here.
Manage AI covers responsible use. Data privacy. Knowing what to share with AI and what to keep private. Understanding bias and how to identify it. Knowing when AI is appropriate and when it is not. This domain also covers the human side: understanding how AI affects wellbeing, attention, and decision-making. Five competences.
Design AI is the most ambitious domain. It asks students to think about the human choices behind AI systems. Who decided what data to use? Who benefits from this AI system? Who is excluded? What would responsible AI design look like? This domain connects directly to ethical reasoning that already appears in subjects like Business Studies, Citizenship, PSHE, and Design and Technology. Four competences.
The 22 competences across these four domains are not a checklist to complete. They are a map of the territory. A school does not need to cover all 22 in a year. It needs to know which competences it is building strength in and where the gaps are.
How AILit connects to what schools already do
The framework was designed to be embedded, not bolted on. The published scenarios show AI literacy activities in primary and secondary classrooms, some of which require no technology at all. The framework document explicitly states that AI literacy can be implemented "in some cases without the need for AI technologies."
This matters because it validates the approach many teachers instinctively take. A History teacher who asks students to evaluate an AI-generated summary is building Engage with AI competences without needing a single laptop. A D&T teacher who asks students to design rules for an AI system is building Design AI competences without writing any code.
The challenge for schools is not understanding the framework. It is mapping the framework to their existing curriculum in a way that is specific enough to be useful. Saying "we develop critical thinking" is not enough. Identifying which specific AILit competences connect to which specific lessons in which specific subjects is the work that makes the framework actionable.
This is the mapping work that AILitKit's Topic guide does automatically. When you upload a scheme of work, the guide analyses every lesson and identifies where each of the four domains connects naturally. The framework alignment section shows which competences the activities develop, labelled with their depth: Introduced, Practised, or Assessed. You do not need to have read the framework document. The guide does the mapping for you.
What PISA 2029 means in practice
Every three years, PISA assesses 15 year olds across OECD member countries. The results shape national education policy for years afterwards. In 2029, PISA adds Media and AI Literacy (MAIL) as a new assessment domain.
Students will be tested on their ability to understand how algorithms work, evaluate AI-generated content for credibility, identify bias in AI outputs, and make ethical decisions about AI use. The test items will be scenario-based. Students will not be asked to define terms. They will be given situations and asked to reason about them.
This is good news for schools that teach through discussion, debate, and evaluation. PISA 2029 will not test whether students can code a neural network. It will test whether students can think critically about one. The teacher who asks "Should we trust this AI's recommendation, and why?" is preparing students for exactly the kind of reasoning PISA will assess.
For UK schools, PISA results influence government policy, Ofsted priorities, and media narratives about school quality. For international schools competing on academic reputation, strong PISA performance is a differentiator. Being able to tell prospective parents "our students are prepared for the PISA 2029 AI literacy assessment" carries weight.
The coverage question governors will ask
At some point, a governor will ask: how does our AI literacy provision map against the frameworks?
This question is reasonable. It is also very difficult to answer without structured tracking. A school that has done good work across six subjects has no way to show that work maps to specific competences unless someone has done the cross-referencing manually. That cross-referencing is time-consuming, requires framework expertise, and changes every time a new guide or activity is added.
AILitKit's coverage heatmap, available on higher tiers, shows the distribution across the four AILit domains at a glance. Green where the school has built strength. Gaps where provision has not yet reached. The governor report translates this into prose that governors can understand: "Coverage is strongest in Engage with AI, with activities across four subjects developing critical evaluation of AI outputs. Design AI remains a gap, with no current activities asking students to consider the human choices behind AI system development."
The heatmap is not about compliance. It is about visibility. A school that can see its coverage can make informed decisions about where to invest next.
The six principles behind AILit
The framework is built on six principles that guide how AI literacy should be taught. Schools do not need to memorise these, but understanding them helps explain why the framework recommends certain approaches.
AI literacy is foundational, meaning every student needs it, not just those in STEM. It is cross-curricular, meaning it should be integrated into existing subjects rather than taught as a standalone course. It is age-appropriate, meaning the same concept should be taught differently at primary and secondary level. It is practice-oriented, meaning students learn by doing, not just by hearing. It promotes agency, meaning students should feel empowered to make decisions about AI, not passive in the face of it. And it requires educator support, meaning teachers need training and resources to deliver it well.
Every one of these principles aligns with how AILitKit works. Cross-curricular by design. Age-appropriate through key stage differentiation. Practice-oriented through activities with scripts, questions, and contingency plans. Agency-building through Designer's Dilemma questions that flip students from users to designers. And educator-supporting through the Preparation section that builds teacher confidence before the lesson begins.
Where to start with AILit
Do not start by reading the 80-page framework document. Start by teaching one lesson with one AI literacy activity and seeing which domain it naturally falls into.
If the activity asks students to evaluate something AI produced, that is Engage with AI. If it asks them to use an AI tool to create something, that is Create with AI. If it asks them to think about privacy, bias, or responsible use, that is Manage AI. If it asks them to think about how an AI system should be designed, that is Design AI.
One activity. One domain. That is how coverage starts. The framework provides the map. Your curriculum provides the territory. AILitKit connects the two.
AILitKit maps every activity to the OECD AILit Framework's four domains automatically. Topic guides show the mapping across a full scheme of work. Whole Curriculum guides show it across subjects and key stages. Try a free Lesson guide at ailitkit.com