| Abstract | Generative AI has created a fundamental crisis in higher education assessment, where traditional output-based evaluation cannot distinguish genuine learner competency from AI-assisted performance. This paper proposes a framework for process-oriented competency evaluation comprising three interconnected components: (1) Competency Model defining what to assess in AI-augmented learning; (2) Analytics Integration capturing processes where learning happens; (3) Actionable Guidelines providing practical tools for educators. Ethical principles are embedded within the framework to ensure responsible practice. Grounded in Self-Regulated Learning, Bloom's Taxonomy, and Learning Design theory, this framework addresses, how authentic learner competency can be evaluated in GenAI-enhanced learning environments, bridging the adoption-to-actionability gap through theoretically grounded, and practically actionable foundations for responsible Generative AI integration that enhances rather than replaces human learning. |
|---|