With the acceleration of urbanization, leftover green spaces in aging communities have long been neglected due to spatial fragmentation and governance absence, highlighting the urgent need for innovative, low-intervention renewal strategies. This study takes the typical fragmented green spaces of Tielu Third Village in Chongqing as a case and conducts a comparative experiment between a “traditional design team” and an “AIGC-assisted design team” to systematically evaluate the applicability of AI-generated content (AIGC) in community green space micro-regeneration. Combining expert reviews, resident interviews, and process documentation, the research compares the two approaches across multiple dimensions, including design efficiency, spatial utilization, ecological performance, innovation, and user acceptance. Results show that AIGC significantly improves design generation speed and visual representation, enhances spatial innovation, and facilitates co-creation with residents. However, it still relies on human intervention for detail refinement, site-specific adaptation, and cultural expression. The study concludes that while AIGC brings efficiency and diversity to green space renewal, its sustainable application depends on strengthening human–AI collaboration and embedding local cultural context. This research provides empirical evidence to support both practice and theoretical advancement in AI-assisted community micro-regeneration.