Designing GenAI Literacy for Cognitive Sovereignty: A GAP-Based Pedagogical Framework
Gábor Soós
Abstract
This chapter develops a GAP-based pedagogical framework for new AI literacy in education. Building on the concepts of the Generative Attunement Point (GAP) and cognitive sovereignty, it argues that the central educational challenge of generative AI is not only whether learners can formulate effective prompts, evaluate outputs, or use tools fluently. The deeper question is whether they can preserve independent reasoning, judgment, and meaning-making in the interval between an AI-generated response and their own next action. Current approaches to GenAI literacy often emphasize prompt design, output evaluation, and tool familiarity. These competencies remain necessary, but they are insufficient if learners accept fluent AI responses without carrying out the cognitive operations through which learning occurs: comparison, synthesis, evaluation, revision, source-grounding, original connection, and accountable judgment. The chapter proposes that the basic pedagogical unit of GenAI literacy is not the prompt alone, but the prompt-response-GAP-re-prompt cycle, understood as a learning interval between machine-generated output and human responsibility. The educational risk is GAP collapse: the narrowing or disappearance of this interval through immediate acceptance, superficial reuse, or passive dependence on AI-generated output. The educational opportunity is GAP expansion: deliberate practice in pausing, verifying, comparing sources, identifying assumptions, formulating counter-questions, revising claims, and deciding what remains the learner's own responsibility. The chapter situates this framework in relation to emerging research on GenAI in education, including evidence that AI can improve task performance while not automatically improving learning, as well as studies on AI-assisted feedback-seeking, student agency, self-regulation, over-reliance, and cognitive laziness. It connects GAP-based pedagogy to intermodal fluency: the capacity to move responsibly among direct human observation, analogue encounter, digital resources, AI-generated synthesis, social dialogue, and accountable human judgment. It also clarifies Co-Prompting as a collaborative human practice across difference rather than as a claim of symmetrical human-machine co-authorship. The chapter concludes with design principles and assessment implications for AI literacy curricula: claim before completion, evidence before fluency, reflection before submission, revision before acceptance, human judgment before automation, and, where collaborative learning is involved, plurality before prompting. It argues that new AI literacy should be understood not as technical competence alone, but as the protection and cultivation of the learning interval in AI-mediated education.