Ruwad Al-Mustaqbal System: A Knowledge-Behavior Gap-Aware Framework (KBG-Ruwad) for Personalized Responsible AI Education among Omani Undergraduates

Abstract

The use of generative AI among undergraduate students has surpassed structured instruction in responsible use of generative AI, and this gap is documented, with students' awareness of and knowledge of AI ethics not matching with their ability to apply it in academic contexts. This has been identified as a critical issue in Oman, with Vision 2040 stating AI adoption as a visionary goal, and most Omani HEIs lack formal AI governance, along with faculty AI literacy falling short of visionary goals. AI ethics are structured into progressive levels in competency frameworks like UNESCO's, and adaptive systems already personalize content. This paper proposes a conceptual framework, KBG-Ruwad, that introduces a novel measure of the knowledge-behavior gap (KBG). The framework assesses, for each student and across each dimension, the discrepancy between self-reported ethical knowledge and behavior demonstrated in scenario-based tasks. In addition, it measures students' level of ethical knowledge, their confidence in that knowledge, and the misconceptions they hold. The framework further extends the existing seven-dimensional model. Building on these measures, KBG-Ruwad extends the seven-dimension Responsible AI Competency Vector with discipline-specific, misconception-aware adaptive scenarios, and with a guidance function sensitive to gap, confidence, discipline and learning history; it further introduces a student-owned Learning Passport and a privacy-protected institutional dashboard. The paper reviews the AI-competency, AI-ethics pedagogy and Omani AI-readiness literature; argues for integration over invention; critically analyzes the components and flow of the proposed framework; and identifies gaps in the current specification and challenges for its operationalization