Autonomous or Static: The effect of AI-enabled environmental benefits in shaping consumer green purchase intention database

This dataset contains three batches of anonymized online scenario survey data of Chinese adult consumers collected from late November to early December 2025, which supports the empirical analysis of the manuscript exploring the dual impact of AI autonomous environmental benefits on consumers’ green purchase intention based on the Theory of Planned Behavior. Three independent between-subjects experiments adopted intelligent air conditioners, eco-backpacks and energy-saving water heaters as distinct product stimuli to avoid single-product bias, respectively verifying the basic mediating path of environmental self-efficacy, the moderating effect of consumer innovativeness and the boundary role of time orientation. We strictly cleaned invalid samples with overly short filling time and identical consecutive answers, retaining 241, 360 and 362 valid respondents for Study 1, Study 2 and Study 3 separately. All latent variables were measured by mature 7-point Likert scales from published literature, covering perceived AI dynamic environmental benefits, environmental self-efficacy, green purchase intention, consumer innovativeness, time orientation, manipulation check items and demographic control variables (gender, age). The archived files include universal CSV and SPSSraw data, as well as an attached code book explaining variable definitions and coding rules. All original statistical results (ANOVA, bootstrap mediation and moderated mediation tests) can be fully replicated using SPSS and Hayes’ PROCESS Macro. No sensitive personal information was collected, all participants provided voluntary informed consent, and the anonymized data is openly available for non-commercial academic replication and secondary research on AI-driven sustainable consumption.