Spanish A parrot in my class: GAI and literacy

Abstract

We examine the integration of Generative Artificial Intelligence (GAI) in three instructional domains: the teaching of writing production, the teaching of reading comprehension, and the use of literacy-based tasks to teach disciplinary content across the curriculum. We assume that GAI is already embedded in school practices, that banning or restricting its use is neither feasible nor pedagogically meaningful, and that educational institutions should instead implement transparency policies that encourage students to acknowledge their use of GAI, scaffold their interaction with it, and promote critical reflection.

In the first domain, writing instruction, we compare the cognitive model of the writing process developed in the 1980s with an analysis of complete GAI-assisted writing logs produced by expert professionals from diverse fields. Our findings reveal substantial transformations in writing practices resulting from ordinary GAI use, including continual prompt reformulation and the disappearance of traditional intermediate products (lists, outlines, drafts). We introduce the concept of analytical reading to describe the complex task of scrutinizing GAI outputs to detect hallucinations, biases, and inconsistencies, and then refining the initial prompt accordingly. Two pedagogical recommendations follow: prioritizing tasks that pose communicative problems without prescribing a specific text type—thereby discouraging simple “copy-and-paste” practices and fostering reflection—and encouraging students to use GAI only after completing their own work, in order to avoid priming effects.

In the second domain, reading instruction, we analyze the strategies through which GAI assists users' comprehension of texts. We propose the concept of mediating artifact to describe these strategies and offer an initial typology based on their linguistic characteristics. This analysis shows that GAI places substantial high-level cognitive demands on users, including technological proficiency, advanced analytical reading skills (which only partially overlap with critical reading), and sustained self-regulation.

In the third domain, literacy across the curriculum, we revisit psychological theories from the 1980s concerning the epistemic nature of writing, which underpin the pedagogical use of literacy tasks for learning disciplinary knowledge. We compare these theories with contemporary GAI-mediated professional practices and, considering their potential classroom applications, hypothesize about the long-term implications for schooling. Our analysis suggests that GAI will profoundly reshape educational practice. By automating written production, shifting cognitive effort toward analytical reading, and substantially reducing learners’ workload, GAI disrupts the pedagogical value of many traditional assignments. Tasks such as summarizing, outlining or preparing written reports become less meaningful for students with unrestricted access to these tools.

Ultimately, the emergence of GAI compels a reevaluation of reading and writing instruction. Process-based writing pedagogy becomes less central, while analytical reading skills gain increasing importance. Other competencies that require renewed emphasis include self-regulation during reading, effective interaction with GAI systems, systematic verification of outputs using reliable sources, the ability to personalize results in one’s own words, and the capacity to articulate and justify viewpoints orally.