Developing and validating a consumer centered VCIO framework for artificial intelligence governance in Indonesia

Artificial intelligence governance frameworks have proliferated globally, yet three critical gaps persist: the principles-to-practice gap, where ethical guidelines remain abstract without measurable implementation mechanisms; the consumer-centric gap, where frameworks are designed primarily from regulator and developer perspectives while consumers account for only 2.2 percent of the actors represented in AI governance scholarship; and the developing country gap, where sophisticated models presuppose institutional capacities absent in emerging economies. This study addresses these gaps by developing and empirically validating a consumer-based AI governance framework using the Value-Criteria-Indicator-Observable (VCIO) model in Indonesia. A sequential mixed-method design was employed, combining confirmatory factor analysis (n = 236) with expert focus group discussion involving three specialists from government, industry, and academia. The CFA confirmed that Sociolegal, Risk, and Functional dimensions constitute valid and reliable constructs, with all indicator loadings exceeding 0.7, construct reliability above 0.88, and discriminant validity established through the Fornell-Larcker and heterotrait-monotrait ratio criteria. The expert phase elaborated the validated structure into eighteen Indicators with concrete Observables and designated measurement approaches, completing the VCIO hierarchy from abstract values to quantifiable data points. The contribution of the study is methodological and operational rather than theory building: to the best of the authors’ knowledge, the resulting framework constitutes the first empirically validated, consumer-centric VCIO instrument for AI governance developed and tested in a developing country, providing an actionable governance tool through which consumer experience can inform regulatory design.