PERSONALIZED MODERATION SHAPES STRUCTURAL ASYMMETRY AND SHARED COGNITION IN ONLINE COMMUNITIES

Social media platforms have recently introduced personalized content moderation (PCM), allowing users to tailor content visibility based on individual preferences. While these tools aim to enhance user experience by aligning content with personal settings, they also pose risks of fragmenting shared informational environments. This paper examines the unintended consequences of PCM on the structural and cognitive dynamics of online communities. Introducing the concept of asymmetric information loss (AIL), we explore how varying PCM settings can fragment shared informational spaces, disrupt semantic alignment, and hinder collective sensemaking. Our computational analysis reveals that greater divergence in PCM settings leads to significant structural AIL, characterized by increased variability in content visibility across users. This fragmentation subsequently weakens shared cognition, particularly in communities focused on fostering debate and diverse perspectives. Our findings emphasize the critical need to balance personalization with the maintenance of cohesive shared contexts, offering insights into the broader implications of PCM on community dynamics and collective understanding.