The 4D Framework for Generative AI Fluency An Instructional Methodology for Higher Education

Abstract

The growing use of generative artificial intelligence (AI) in higher education demands the development of competencies that go beyond the instrumental use of specific tools. This article describes the design and theoretical foundations of an instructional methodology for the transversal integration of AI fluency in higher education programs, based on the 4D Framework (Delegation, Description, Discernment, and Diligence) developed by Dakan and Feller (2024) in collaboration with Anthropic. The methodology proposes four progressive levels of curricular integration—Exploration, Application, Integration, and Mastery—articulated with three human-AI interaction modes: automation, augmentation, and agency. The article presents the instructional design by competency, ten conceptually validated learning activities for five disciplinary areas, an integrated assessment rubric, and a four-phase institutional implementation pathway. This proposal represents a novel contribution in the Chilean and Ibero-American context, and aims to guide higher education institutions in developing transferable, durable, and ethically grounded AI competencies in their students.