From Detection to Prevention: A Human-Centered Design ‎‎(HCD) Approach to Mitigating AI Misuse in Learning

The rapid integration of artificial intelligence (AI) tools in educational settings has created unprecedented challenges for academic integrity, ‎with studies indicating that 89% of students are familiar with ChatGPT and 43% have used AI tools for academic work, while current detection-based approaches have proven inadequate, achieving only 26% accuracy in identifying AI-generated content. This study develops and ‎validates a comprehensive Human-Centered Design (HCD) framework for mitigating AI misuse in educational contexts, shifting from reactive detection strategies to proactive prevention-oriented approaches that prioritize stakeholder engagement and ethical AI integration. Using ‎a two-phase mixed-methods approach combining Nominal Group Technique (NGT) and Fuzzy Delphi Method (FDM) with seven experts, ‎the research achieved exceptional expert consensus with an overall average agreement of 98% across all evaluated components. Eleven core ‎elements were validated and organized into five interconnected dimensions: Core Design Principles (emphasizing learner agency and educator empowerment), Implementation Strategies (featuring progressive disclosure and reflection mechanisms), Assessment & Feedback Sys‎tems (incorporating authentic assessment and ethical analytics), Stakeholder Engagement (encompassing co-design processes), and Ethical ‎Safeguards (focusing on bias detection and privacy protection). Four components achieved perfect 100% expert consensus, while the low-‎lowest-ranked element maintained an acceptable defuzzification value of 0.800, well above the 0.5 threshold required for validation. The validat‎ed HCD framework provides a robust, empirically supported approach for educational institutions to proactively address AI misuse while ‎maintaining the benefits of AI integration, offering a paradigm shift from punitive detection methods to collaborative, prevention-focused ‎strategies that enhance rather than restrict educational AI applications, thereby contributing to the emerging fields of Human-Centered Learn‎ing Analytics (HCLA) and Human-Centered Artificial Intelligence (HCAI) with practical guidance for educational stakeholders seeking ‎responsible AI integration strategies.