Developing AI literacy

Exposure is not development

Most people who use AI regularly don’t become AI literate through use alone. Exposure produces familiarity, the ability to get something useful out of an AI interaction. Development produces capability, the ability to judge when and how using AI serves the goals you actually have, and it takes deliberate practice and reflection to get there.

Developing AI literacy

One-sentence definition: Developing AI literacy is the gradual deepening of how a person works with AI, from substitution through adaptation to transformation, during which the six dimensions of AI literacy grow together through real use and deliberate reflection.

A person can become highly familiar with AI tools, finding their way around the interfaces, getting reasonable outputs, and avoiding obvious errors, without developing the contextual judgement, critical evaluation, or ethical awareness that make up literacy. The difference matters because familiarity is comfortable and reinforces itself: outputs improve, which feels like progress, while the deeper dimensions remain undeveloped.

The three-stage trajectory

The developmental stages reflect a recognisable pattern in how people integrate new capabilities into professional practice.

Substitution¹ is the natural entry point because cognitive load is lowest: you’re using AI for tasks you already know how to do, so you can check the outputs against your own knowledge. This stage mainly develops functional application and surface-level critical evaluation. It’s a useful place to start, but it tends to limit itself, because once a workflow is efficient there’s little internal pressure to go further.

Adaptation means letting AI change how you approach a task. Instead of using AI where you’d have used Google or a blank page, you restructure the activity around what AI makes possible: building understanding through staged dialogue, testing arguments by having them questioned, and breaking a problem down before trying to put it back together. This stage develops creation and communication and deepens access and recognition.

Transformation can’t be designed or rushed. It emerges from sustained practice, accumulated context, and growing professional judgement, as AI becomes something you think with and alongside, and what comes out of that thinking informs other parts of your practice and learning. Context builds across sessions, the ways you use AI become deliberate and specific to your field, and the central question shifts from “how do I use this?” to “when does this serve my goals?” This is where contextual judgement and ethical awareness develop most fully.

Take a student radiographer writing reflective accounts on placement. At substitution, they write the reflection and ask AI to tidy the prose, and they can judge the result because they know what they meant. At adaptation, they talk through a positioning decision with AI before writing, using it to test their reasoning about why an image came out as it did.

At transformation, which some students reach only late in a programme, AI has become something they think with and alongside. Their reflections from earlier placements are a record they work through with it, looking for patterns no single account shows (the same positioning error with larger patients, the same feedback from different educators), and what they find shapes what they bring to supervision and the goals they set for the next placement. They still write each reflection themselves, because the account exists to do that thinking; AI’s part is the view across time that informs everything around it.

What developing AI literacy requires

Three conditions separate development from exposure.

Foundational orientation. Before development can begin, learners need to understand what AI systems are, how they generate responses, and where they fail. They don’t need to study AI as a topic, only enough of how it works to make critical evaluation possible. Without that grounding, the first stage collapses into uncritical use.

Authentic use. Literacy develops through real tasks with real stakes. The disciplinary or professional content comes first, and AI literacy develops as a by-product of working on it; exercises that practise AI use for its own sake do little.

Deliberate reflection. Moving between stages takes metacognitive work: evaluating how you worked with AI as well as what it produced. Without it, people stay at substitution. Reflection surfaces the dimensions that are not developing and creates the conditions for progression. Progression usually needs some disruption: a task that substitution handles poorly, or a reflective prompt that exposes the gap between familiarity and competence.

See AI literacy development framework for practical guidance on embedding these conditions into existing modules and courses.

Sources

• Long, D., & Magerko, B. (2020). What is AI literacy? Competencies and design considerations. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems (pp. 1–16). ACM. https://doi.org/10.1145/3313831.3376727
• Miao, F., & Holmes, W. (2023). Guidance for generative AI in education and research. UNESCO. https://doi.org/10.54675/EWZM9535
• Mollick, E. R., & Mollick, L. (2023). Assigning AI: Seven approaches for students, with prompts [Preprint]. SSRN. https://doi.org/10.2139/ssrn.4475995