<p>As a subset of artificial intelligence (AI), generative AI (GenAI) is evolving rapidly and has attracted increasing attention in technology and education research. At the same time, digital media platforms such as Douyin have seen a growing amount of user-generated micro-learning content related to GenAI. However, the extent to which such content stimulates GenAI adoption, and the conditions under which adoption may be constrained, remain insufficiently understood. Drawing on the technology acceptance model, this study examines how exposure to GenAI-related micro-learning content influences technology acceptance among college students, while highlighting the role of AI identity threat as a moderator. An online survey was conducted with 513 students majoring in music-related fields from three universities in China. The results reveal that higher perceived quality of micro-learning content is positively associated with perceived usefulness and perceived ease of use of GenAI. These perceptions contribute to more favorable attitudes toward GenAI, which in turn predict stronger behavioral intentions to use and recommend the technology. Importantly, AI identity threat emerges as a significant limiting factor. It is directly and negatively associated with behavioral intention to adopt GenAI and moderates the attitude-behavior relationship, curtailing the translation of positive attitudes into acceptance intentions. This suggests that even when GenAI is perceived as useful and easy to use, concerns about identity threat can substantially hinder its adoption. Overall, this study extends technology acceptance research by integrating micro-learning and identity-based social concerns, generating insights relevant to educational and broader technology adoption contexts.</p>