Despite the widespread acceptance of sustainable development principles, effectively engaging consumers remains a core challenge for businesses and policymakers. The emergence of generative artificial intelligence (GAI) presents both new opportunities and exacerbates uncertainty in the information environment. To provide practical insights, this study employs a comprehensive approach combining quantitative and predictive analytics within the framework of the Theory of Planned Behavior (TPB). This research utilizes structural equation modeling, cluster analysis, and Bayesian networks to systematically validate the relationships between key psychological variables and simulate behavioral intervention pathways within the TPB framework. The study finds that: 1) the diversity of GAI information lacking transparency weakens consumers' perceived behavioral control by exacerbating uncertainty, thereby reducing their sustainability focus and perceived value. 2) Consumers with a high sense of responsibility exhibit stronger resilience and behavioral conversion potential, while cost sensitivity significantly modulates the path from attitude to behavioral intention. This paper provides a theoretical basis and an actionable framework for businesses to effectively guide green consumption behavior by building a trustworthy information environment, implementing precise communication, and optimizing cost perception.