Assessment by design

A classification framework for learning assurance in the age of GenAI

Abstract

Assessment classification plays a critical role in ensuring transparency, coherence, and academic integrity across higher education curricula. The rapid emergence of generative artificial intelligence (genAI) has intensified the need for institutions to assure learning outcomes that are demonstrably independent of AI-generated content. This challenge is compounded by the need to manage institutional risks, particularly misconduct arising from inappropriate genAI use, and to uphold assessment validity as a measure of student learning.

Despite this urgency, there remains a lack of structured, scalable frameworks that support both pedagogical intent and institutional governance in assessment classification. This paper addresses that gap by proposing a revised classification system, comprising assessment categories and types, designed to meet the diverse needs of students, educators, and governance bodies. The model aims to enhance clarity, support constructive alignment, and enable data-informed decision-making across the curriculum.

Drawing on TEQSA’s genAI principles and the Australian Qualifications Framework (AQF), the approach integrates verb-based classification aligned with learning outcomes to support a course-level approach to assessment. While implementation is ongoing, the proposed model offers a practical, standards-aligned response to institutional and regulatory challenges, contributing a framework that supports a pedagogically sound assessment while strengthening institutional capacity to assess and mitigate risk.