Understanding Safety Risks and Safety Design in Social VR Environments

Understanding the emerging safety risks in nuanced social VR spaces and how existing safety features are used is crucial for building safe and inclusive 3D virtual worlds in the future. Prior research on safety risks in social VR is mainly based on interview or survey data of social VR users' experiences and perceptions, which lacks "in-situ observations'' about how people actually react to these risks. In this paper, we aim to understand safety risks and safety design in social VR environments using two empirical studies. In Study 1, we conducted content analysis of 212 YouTube videos and the associated transcripts that document social VR users' first hand experiences of safety risks in social VR environments in the moment (e.g., as victims, attackers, or bystanders). We also investigated viewers' reactions to these risks by analyzing comments to these videos posted online (i.e., spectators). In Study 2, we identified 13 safety features across various social VR platforms and mapped how each existing safety feature in social VR can address the risks identified in Study 1. Our findings call for re-approaching safety risks based on the uniqueness of social VR interaction dynamics and users' multi-modal simulated reactions to such risks. We also propose three safety protection mechanisms for designing future social VR environments: the use of non-player character (NPC) as a safety educator, intuitive gesture-based safety triggers, and improvements for avatar and voice control.