AI literacy is a term almost everyone in education uses, and, as Yu (2025) shows, almost no one defines the same way. Ask three researchers what it means and you get three answers: knowing what AI is, using AI tools, or thinking ethically about AI. Yu sets out to bring order to that sprawl by viewing the term through a single lens, competency.
Yu concludes that “the previous research on the AI literacy that learners develop and learn through AI education is fragmented and poorly integrated” (p. 2990). To map the fragmentation, Yu examines 16 studies published between 2012 and 2023 and sorts them into four groups: studies that break AI literacy into components, studies that tie it to computational thinking, studies that ground it in competence, and studies that treat ethics as a separate track.
The Same Three Elements Across the Field
Across all four groups, Yu reports, the same three elements keep appearing: knowledge, ethics, and problem-solving. The studies rarely build on one another, which is how the field ends up with a scattered picture of a single idea. Long and Magerko’s four questions (cited in Yu, 2025) organize the knowledge side. Wang and colleagues (cited in Yu) add evaluation and ethics to a measurement scale. Holmes and colleagues (cited in Yu) treat ethics as its own six-area domain.
Yu then separates two terms that often get used interchangeably. In Armstrong’s terms, competence is the functional ability to perform a task, and competency is the behavioral dimension underneath that performance. Because most AI literacy research focuses on what learners can produce, Yu explains, “this article uses the term competency rather than competence” (p. 2984).
On that basis, Yu adopts the Competency Learning Framework (CoLeaF) from Frezza and colleagues (cited in Yu), which has four parts: Knowledge, Skills, Dispositions, and Context. Yu maps the existing studies onto these parts. Component and computational-thinking work feeds Knowledge. Studies on using, applying, evaluating, or creating with AI map to Skills. The ethical and affective research fills Dispositions. The competence-perspective studies become Context, the authentic situation in which the others operate.
A Non-Linear Model
Yu first sketched the framework as a straight line, then revised it. The final version places Context as a large circle around Knowledge, Skills, and Dispositions, with dotted, overlapping borders to show that the parts blend into one another. Yu states that “since the components of AI literacy in the suggested framework are interrelated, it does not represent sequential learning” (p. 2989). A learner does not complete the knowledge stage before moving to skills.
Yu also acknowledges a tension in the project. A paper that criticizes a crowded field then adds one more framework to it. Yu presents CoLeaF as a consolidating lens over existing work, not a new rival. The fragmentation diagnosis lines up with the rapid review from Hillman and colleagues (2025), which found AI literacy frameworks plentiful but short on shared measures. The competency focus connects to Chee and colleagues (2025) on competency mapping for different learner groups, and to Chiu (2025) on the definitions of literacy and competency.
Scope and Limits
Yu is clear that this is a conceptual paper with no empirical test behind it, and presents the framework as a reference for future studies, including work on measuring and scaling AI literacy. The paper positions AI literacy as a basic literacy for every learner, not a computer-science specialty, and argues that a holistic, competency-based view suits K-12 curriculum design better than a checklist of technical skills. Yu offers the four-part structure, knowledge, skills, dispositions, and context, as a shared starting vocabulary for AI literacy.
References
- Chee, H., Ahn, S., & Lee, J. (2025). A competency framework for AI literacy: Variations by different learner groups and an implied learning pathway. British Journal of Educational Technology, 56, 2146-2182. https://doi.org/10.1111/bjet.13556
- Hillman, V., Holmes, W., & Duarte, T. (2025). A rapid review of AI literacy frameworks, with policy recommendations. A report prepared for the Royal Society. London: The Royal Society. https://royalsociety.org/-/media/policy/projects/ai-in-education/hillman-et-al-a-rapid-review-of-ai-literacy-frameworks.pdf
- Yu, W. (2025). A conceptual framework for AI literacy with a focus on competency. International Journal of Artificial Intelligence in Education, 35(4), 2975–2992. https://doi.org/10.1007/s40593-025-00488-4