Hybrid Model for Analyzing Consumer Adoption Decisions Regarding Generative AI: An ExtendedTAM-Based Framework

In this study, a hybrid multi-criteria decision-making (MCDM) model was developed for analyzing consumer adoption decisions regarding generative artificial intelligence (Gen AI). By extending the technology acceptance model (TAM) into a structured decision system, the proposed framework integrates ethical and risk-related criteria, including perceived cost, perceived risk, transparency, accountability, intellectual property concerns, and data privacy, into a formal causal and evaluative structure. First, a Delphi-based consensus process is employed to identify and refine key adoption criteria. Subsequently, the decision-making trial and evaluation laboratory (DEMATEL) method is applied to quantify causal relationships among these criteria and to construct an influence network revealing prominence and directional effects. In total, 251 questionnaires were distributed in Taiwan, and 231 valid responses were collected. The results indicated the decision-making factors that underlie the adoption of Gen AI by consumers. The results highlighted transparency as a dominant causal factor that significantly influences multiple ethical and functional dimensions of Gen AI adoption. To address uncertainty and vagueness in human judgment, fuzzy importance–performance analysis was also incorporated. Best non-fuzzy performance values were obtained through defuzzification, enabling the classification and prioritization of critical adoption factors within a four-quadrant decision matrix. The proposed framework provides a mathematically grounded decision-support model for elucidating the structural interdependencies among adoption criteria and to facilitate strategic decision making for Gen AI system design and governance. This study contributes to the MCDM and operations research literature by transforming a behavioral acceptance model into a formal decision-analytic framework, thereby enhancing the analytical rigor and applicability of TAM-based adoption studies in complex socio-technical systems.