Artificial intelligence (AI) technology can efficiently produce media content, including highly realistic images of natural scenery, which holds significant implications for advertising and marketing, particularly in the realm of environmental appeals. However, the uncritical use of AI-generated images may also elicit unintended negative consequences. This study aims to examine the psychological mechanisms explaining consumer responses to the disclosure of AI-generated content in environmental marketing. Specifically, we propose that when consumers identify a natural scene as AI-generated, their willingness to engage in pro-environmental behavior diminishes, and this effect is mediated by a diminished sense of awe. Furthermore, this negative impact is moderated by the level of inclusion of nature in the self (INS) and familiarity with generative AI technology of consumers. Three experiments (total N = 670) were conducted to examine proposed hypotheses. The experimental design manipulated AI disclosure by altering whether the natural landscape images were labeled as "generated by AI". Consumers' sense of awe, pro-environmental behavior intentions, INS, and familiarity with generative AI were measured with established psychological scales. The results of ANOVA and bootstrapping analyses support our hypotheses. The findings reveal that disclosing AI usage can decrease consumers' sense of awe, reducing their intention to engage in pro-environmental behavior. However, consumers with higher levels of INS or greater familiarity with generative AI technology are less susceptible to these negative outcomes. For marketing practitioners, these findings underscore the importance of carefully considering consumers' psychological traits, such as their connection to nature and familiarity with AI, to enhance the effectiveness of AI-generated imagery in campaigns.