Digital competency frameworks in the personalization of artificial intelligence

Abstract

As artificial intelligence (AI) models increasingly play a critical role in society, the demand for AI personalization is becoming a dominant and irreversible trend. This paper analyzes the feasibility and necessity of user-directed AI training and identifies the essential digital competencies required by individuals and organizations to implement personalization effectively and ethically. Based on a literature review, user survey, experiments with LLaMA 3 and GPT-4-turbo models, and a cost-effectiveness analysis, the study demonstrates that personalized AI training is both technically viable and a means to enhance AI democratization. The paper proposes a digital competency framework to support proactive, transparent, and sustainable engagement with AI.