ORCID

Abstract

AI literacy can be situated at the intersection of other types of literacy such as media literacy anddigital literacy, but the concept is distinct. What complicates our understanding of AI literacy is thatAI is highly opaque and black-box, creating obstacles to grasping what these systems are doing andhow (Burrell, 2016; Petrovčič, et al. 2024). This requires an interdisciplinary effort in conceptualizingand mapping AI literacy. Based on our analyses of the current literature, we see five different typesof conceptualizations of AI literacy:1.Competency-based approaches: focused on skills, knowledge, and ethical use of AI.2.Awareness-oriented approaches: centered on recognizing the presence and influence of AI oralgorithms on everyday life.3.Technical approaches: emphasizing programming, system design, or operational understandings.4.Critical or sociotechnical approaches: focused on power, inequality, governance, and resistance.5.Specialized frameworks: tailored to particular professions, sectors, or use cases.We also identified several limitations. Often these approaches make the assumption that individualsare voluntarily engaging with AI rather than being subjected to AI. We found that the majority ofstudies on AI literacy build on this assumption and aim to strengthen individual skill development,moving attention away from collective resistance strategies, structural reforms, and institutionalaccountability. There is a need for more studies that conceptualize AI literacy based on the critical orsociotechnical approaches and we stress the need for critical AI literacy. Research needs to continueexamining how AI literacy is conceptualized across disciplines, regions, and stakeholders, as how wedefine AI literacy influences what individuals and communities learn and how they can engage withAI systems locally and globally.Furthermore, we found that how AI is conceptualized directly shapes how AI literacy is understood.Definitions that focus narrowly on the technical aspects of AI risk overlooking the broader systemicinfluences on AI development, deployment, governance, and experience. Future studies shouldtherefore adopt the term AI ecosystem, rather than relying on use of the terms AI or AI technologies.This language emphasizes the sociotechnical nature of AI, seeing AI as interactively shaped bysymbiotic social, economic, political, and technological relationships (Livingstone, 2004; Pedreschi,2023).

Publication Date

2026-06-05

Publisher

ZENODO

Deposit Date

2026-08-24

Keywords

critical thinking, artificial intelligence, social inequality, information literacy, Educational Technology, community integration, artificial intelligence/trends, Educational Technology/education

Creative Commons License

Creative Commons Attribution-Share Alike 4.0 International License
This work is licensed under a Creative Commons Attribution-Share Alike 4.0 International License.

Download

Share

COinS