AI Literacy Is Not Digital Literacy: Why Schools Need to Stop Conflating the Two

There is a phrase I hear in almost every school leadership meeting about AI: "We already cover this in our digital literacy programme." It is said with confidence, usually accompanied by a reference to the school's existing e-safety curriculum, its online citizenship module, or its ICT programme. The implication is that AI literacy is simply the next chapter of digital literacy; a natural extension of what schools are already doing.

It is not. AI literacy and digital literacy are different things. They require different knowledge, different skills, and different dispositions. And schools that treat AI literacy as a subset of digital literacy are leaving their students and staff fundamentally unprepared for the world they are entering.

What Digital Literacy Is

Digital literacy, in its most widely used formulation, encompasses the skills needed to effectively and safely use digital technologies. It includes practical skills (using software, navigating the internet, managing files), information skills (searching, evaluating sources, identifying misinformation), communication skills (email etiquette, online collaboration, social media awareness), and safety skills (passwords, privacy settings, recognising scams, reporting abuse).

These are important skills. Schools should continue teaching them. But they describe a relationship with technology that is fundamentally different from the relationship students have with AI.

Digital literacy assumes that technology is a tool; something the user controls. You open a browser, you type a search query, you evaluate the results, you decide what to do with the information. The technology responds to your commands. You are the agent; the technology is the instrument.

What AI Literacy Is

AI literacy describes a fundamentally different relationship. AI systems are not just tools that respond to commands; they are systems that generate, interpret, predict, and sometimes decide. When a student interacts with an AI chatbot, the system is not passively waiting for instructions. It is generating content, making choices about what to include and exclude, and presenting information in a form that reflects its training data, its design constraints, and its probabilistic reasoning.

The student is still an agent, but they are interacting with something that has a degree of agency itself, or at least a convincing appearance of it. This changes everything about what "literacy" means.

AI literacy includes understanding what AI is and how it works at a conceptual level: not the mathematics, but the principles. Students need to know that a language model generates text by predicting the most probable next word, not by understanding meaning. They need to know that AI "confidence" is statistical, not epistemological. They need to understand that AI systems reflect the data they were trained on, including its biases, gaps, and cultural assumptions.

AI literacy includes the ability to critically evaluate AI output, not just for factual accuracy (which is part of digital literacy's information evaluation skills) but for bias, framing, perspective, and the subtle ways in which AI-generated content shapes understanding without the user realising it. A student who reads a Google search result and evaluates its reliability is exercising digital literacy. A student who recognises that an AI-generated summary of a historical event reflects a particular cultural perspective and can articulate why is exercising AI literacy.

AI literacy includes understanding the ethical dimensions of AI: privacy, consent, fairness, accountability, transparency. These overlap with digital literacy's privacy and safety skills, but they go much deeper. Who is responsible when an AI system makes a wrong recommendation? What does it mean to consent to your data being used to train a model? Is it fair for an AI to make decisions about students based on patterns in historical data that may reflect systemic inequalities? These are not safety questions; they are philosophical and ethical questions that require a different kind of engagement.

AI literacy includes the ability to use AI as a thinking partner rather than an answer machine. This is a skill that has no equivalent in digital literacy. Digital literacy teaches you to find information. AI literacy teaches you to generate, evaluate, refine, and build on ideas in collaboration with a system that can process and produce language at scale. It is a creative and critical skill, not just an information management skill.

Why the Conflation Is Harmful

When schools subsume AI literacy under digital literacy, three things go wrong.

First, AI-specific knowledge gets lost. The conceptual understanding of how AI works, probabilistic generation, training data, model limitations, is not covered by any existing digital literacy framework. If it is not taught explicitly, students will form their own mental models, and those mental models will be wrong. Students who believe AI "understands" their questions, "knows" the answer, or is "thinking" about their problem will use AI tools in fundamentally flawed ways.

Second, the ethical dimension gets flattened. Digital literacy ethics centres on personal safety and responsible online behaviour; do not share your password, do not bully people online, be careful what you post. These are important but they are not the ethics of AI. AI ethics involves questions about algorithmic fairness, institutional decision-making, data consent at scale, and the societal implications of automating human judgement. Treating AI ethics as an extension of e-safety trivialises it.

Third, the critical evaluation skills stay too shallow. Digital literacy teaches students to evaluate sources: Is this website reliable? Does this article cite evidence? Is this news or opinion? These skills transfer to AI, but they are not sufficient. AI does not produce "sources" in the traditional sense; it generates content that is not attributable to any specific source, that may blend accurate and inaccurate information seamlessly, and that presents everything with equal confidence regardless of reliability. Evaluating AI output requires a different skill set: understanding when and why AI is likely to be wrong, recognising hallucination, assessing whether the AI's framing serves the user's purpose, and knowing when to trust and when to verify.

What Schools Should Do

Recognise AI literacy as a distinct domain. It is not digital literacy plus a module on ChatGPT. It is a separate set of competencies that requires its own curriculum time, its own learning outcomes, and its own assessment.

Map the overlap and the gaps. There is genuine overlap between digital literacy and AI literacy; both involve critical evaluation, both involve ethical reasoning, both involve safe and responsible use of technology. Identify where your existing digital literacy provision covers AI-relevant skills and where it does not. The gaps are your AI literacy curriculum.

Start with what AI is, not what AI does. Most schools introduce AI through tools, "here is ChatGPT, here is how to use it." This is the equivalent of teaching digital literacy by showing students how to use Google without explaining what the internet is. Start with conceptual understanding: how do language models work? What is training data? What does "probabilistic" mean in this context? Why does AI sometimes produce confident-sounding nonsense?

Embed AI literacy across the curriculum, not just in ICT. Digital literacy made the same mistake for years, confining it to the ICT department rather than embedding it across subjects. AI literacy is relevant in every subject. English teachers can explore AI-generated text and authorship. Science teachers can examine AI's role in research and its limitations in scientific reasoning. History teachers can analyse how AI represents historical events and whose perspectives are centred or marginalised. Art teachers can engage with questions of creativity and originality. AI literacy is not a technology subject, it is a cross-curricular competency.

Assess it. You cannot develop what you do not measure. Build AI literacy outcomes into your curriculum and assess students against them. The AI Literacy Audit Tool assesses the student AI literacy dimension of your school's readiness, benchmarking it against 33 international frameworks. Use it, or develop your own assessment, to track whether your students are actually developing the AI-specific competencies they need.

The Bigger Picture

The distinction between AI literacy and digital literacy is not a semantic quibble. It reflects a genuine shift in the relationship between humans and technology. For thirty years, digital literacy has served us well; it has equipped students to use, navigate, and evaluate the digital world. But the digital world has changed. It is no longer just a repository of information to be searched and evaluated. It is now a generative environment where AI creates content, makes recommendations, and influences decisions at scale.

Students who enter this world with only digital literacy skills are like drivers who know how to use a manual car being handed the keys to an autonomous vehicle. The basic principles transfer; you still need to understand roads, destinations, and safety. But the nature of the machine has changed, and the skills needed to interact with it effectively and critically have changed with it.

Schools that recognise this shift and build AI literacy as a distinct, rigorous, cross-curricular competency will prepare their students for the world as it is becoming. Schools that treat it as a footnote to their existing digital literacy programme will not. The distinction matters, and the time to act on it is now.