From Product to Process: A Framework and Practical Toolkit for AI-Aware University Assessment

The routine availability of generative artificial intelligence (GenAI) has weakened the validity assumptions behind many university assessment formats, especially those that rely on a polished final product as the main evidence of learning. This paper develops a discipline-neutral, process-based framework for AI-aware assessment that treats GenAI as a persistent feature of students’ academic work rather than as a temporary anomaly. Drawing on recent scholarship on assessment validity, authenticity, transparency, and institutional governance, the paper synthesizes key risks of product-only assessment and translates them into a practical redesign logic. The framework rests on five principles: transparency of AI use, auditability through evidence-of-work, visibility of reasoning, contextual authenticity, and feasibility with inclusion. Building on these principles, the article proposes a toolkit that combines staged submissions, annotated decision logs, targeted checkpoints, short oral validations, and verification-by-design in quantitative and applied tasks. Three higher-education illustrations are used to show how the framework can be adapted to writing-intensive, quantitative, and team-based assignments. The paper argues that the central assessment question in the GenAI era should shift from authorship policing to competence verification, and it offers reusable design structures, rubric dimensions, and implementation guidance for institutions seeking to preserve academic standards while supporting responsible AI use.