earticle

영어

The rapid expansion of AI-generated content (AIGC) has raised concerns regarding authenticity and content quality, prompting platforms and regulators to introduce regulations. While these measures are designed to preserve the integrity of online discourse, they may also create unintended participation costs that discourage user engagement and reshape community norms in unexpected ways. This study examines how AIGC regulations shape community dynamics on Reddit, focusing on user engagement and content diversity. Especially, we focus on both text and image content, exploring the multimodal content outcomes. Drawing on signaling theory, we posit that these rules function both as constraints and as authenticity signals that reshape user behavior across community types. Using a quasi-experimental design, our findings reveal that AIGC restrictions reduce participation and diversity in information-focused communities yet stimulate engagement and broaden diversity in creativity-oriented spaces where originality is central. These findings highlight the divergent consequences of identical policies, emphasizing the need for nuanced moderation strategies, underscoring the importance of designing tailored moderation approaches that recognize community heterogeneity. Theoretically, we extend signaling theory into the domain of platform governance, illustrating how authenticity signals shape participation in digital environments. Practically, our findings provide actionable insights for platform operators and policymakers navigatingthe complex challenges of generative AI.