When AI summarizes the community: how generative AI summaries shape users’ intention to engage with original content

The integration of Generative Artificial Intelligence (GenAI) into user-generated content (UGC) communities has fundamentally reshaped how users seek and consume information. Many UGC platforms, such as RedNote and Zhihu, now incorporate GenAI summaries that provide concise, AI-generated answers at the top of search results. While these summaries improve efficiency, they may also reduce user attention to original community posts, thereby threatening engagement and participation within UGC ecosystems. This study aims to examine how GenAI summary design influences users’ intention to engage with community content and the underlying psychological mechanisms. Drawing on Socio-Technical Systems Theory (STS), this research develops a socio-cognitive model linking AI summary design (source localness and source transparency) to engagement intentions through collective psychological ownership and interest-type curiosity. Two online experiments were conducted: Experiment 1 tested the main and mediating effects, and Experiment 2 introduced query type (subjective-experiential vs objective-analytic) as a contextual moderator. Results reveal two socio-cognitive pathways linking GenAI summary design to user engagement. Community-focused and aggregate-list summaries enhance engagement via collective ownership, whereas web-wide and item-level citation summaries foster curiosity-driven engagement. These effects are contingent on query type: subjective-experiential queries strengthen the ownership pathway, while objective-analytic queries amplify the curiosity pathway. This study extends STS to AI-mediated content ecosystems by conceptualizing GenAI summaries as socio-technical interfaces. It advances our understanding of how interface-level design transforms technological affordances into social and cognitive motivations that sustain participation in digital communities.