Echoes of Identity: The Silent Privacy Threats of VR Behavioral Tracking

Abstract

This study examines privacy risks linked to the use of behavioral data for user identification in immersive virtual reality (VR) applications. With advancements in VR technology, tracking sensors now provide highly immersive experiences that capture extensive and nuanced behavioral data. However, limited research addresses the privacy implications of this data collection. In this work, we investigate the potential for machine learning algorithms to identify VR users across multiple sessions and activities and assess their effectiveness even when users alter their behavior to evade detection. Additionally, we explore how physical characteristics impact identification accuracy. Our findings reveal that users can be identified with 83% accuracy across repeated sessions of the same activity and 80% accuracy when performing different tasks, while attempts to mask behaviors still result in 78% accuracy. These results underscore the necessity for enhanced privacy measures to protect user behavior data in VR environments.