Scaling AI Literacy: A Design Framework for University Assessment Alignment

Abstract

Universities are currently transitioning from ad hoc AI tool tips toward institutional strategies, yet they face a significant bottleneck in the absence of scalable, curriculum-embedded AI literacy. This research addresses the need for a coherent, ethically grounded, and assessable framework to integrate generative AI into higher education. The study aims to propose an "AI-literacy-at-scale" model that aligns global UNESCO competency frameworks with institutional curriculum design. Using an integrative synthesis approach, the research analyzes global frameworks, policy guidance, and recent evidence of generative AI adoption. The methodology involves extracting competency descriptors, mapping them to constructive alignment principles, and triangulating these findings with sector-wide governance standards. The study focuses on deriving design principles for outcomes, staff capability, assessment redesign, and quality assurance. The principal result is an alignment matrix and a set of rubric-ready learning outcomes that are adaptable across various academic disciplines. The major conclusion is that embedding AI literacy as a durable graduate capability requires a whole-of-institution approach to safeguard human agency and academic standards. This work contributes a practical blueprint for universities to move beyond tool-centric training toward systemic, ethically grounded curriculum integration.