Generative Artificial Intelligence, Assessment Redesign, and Cognitive Apprenticeship in Higher Education: A Comparative Case-Based Analysis of Process-Oriented and Product-Oriented Learning Governance

Abstract

This article examines how generative artificial intelligence reshapes assessment, student cognition, academic integrity, and pedagogical governance in higher education. While universities increasingly respond to AI through detection systems, policy statements, and academic misconduct procedures, learning sciences research suggests that assessment transformation must move beyond surveillance toward cognitive apprenticeship, reflective learning, and process-oriented evidence of understanding. Using a comparative case-based mixed-methods design, this study analyzes two institutional assessment models: a product-oriented AI-control model and a process-oriented AI-integrative model. The analysis draws on institutional policy documents, assessment rubrics, learning analytics, student performance indicators, classroom observations, and international education reports from 2023–2026. The findings indicate that product-oriented assessment governance may preserve procedural compliance but often weakens cognitive transparency and student trust. In contrast, process-oriented AI-integrative assessment strengthens metacognitive engagement, disciplinary reasoning, collaborative learning, and academic resilience when supported by explicit pedagogical scaffolding. The study argues that generative AI does not simply threaten academic integrity; rather, it exposes deeper weaknesses in assessment systems that privilege final products over visible learning processes. This article contributes to learning sciences scholarship by proposing an assessment transformation framework linking AI literacy, cognitive apprenticeship, process evidence, feedback dialogue, and educational trust.