VR ProfiLens : User Profiling Risks in Consumer Virtual Reality Apps

Ismat Jarin 2 4 , Olivia Figueira 2 , Yu Duan 2 , Tu Le 3 and Athina Markopoulou 2 · 2026

Read VR ProfiLens : User Profiling Risks in Consumer Virtual Reality Apps by Ismat Jarin 2 4 , Olivia Figueira 2 , Yu Duan 2 , Tu Le 3 and Athina Markopoulou 2 (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-12563.txt.

Introduction

Ismat Jarin, Olivia Figueira, Yu Duan, Tu Le, and Athina Markopoulou study privacy risks in consumer virtual reality apps. VR platforms and apps gather rich sensor data from users. The data can cover motion, face, eyes, and hands, though often in abstracted form. Even abstract data may expose users to profiling risks they do not know about.

The team present a tool and study named VR ProfiLens. It probes how apps could build user profiles from such sensor streams. The work was set for a 2026 symposium on usable security and privacy. Findings here matter as VR moves into homes and schools. Users deserve to know what their movements can reveal about them.

Catalog record

CREATORIsmat Jarin 2 4 , Olivia Figueira 2 , Yu Duan 2 , Tu Le 3 and Athina Markopoulou 2
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE114 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-12563

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

VR ProfiLens : User Profiling Risks in Consumer Virtual Reality Apps PubID: pubid: Symposium on Usable Security and Privacy (USEC) 2026 27 February 2026, San Diego, CA, USA ISBN 978-1-970672-07-7 https://dx.doi.org/10.14722/usec.2026.23003 www.ndss-symposium.org, https://www.usablesecurity.net/USEC/ Ismat Jarin 2 4 , Olivia Figueira 2 , Yu Duan 2 , Tu Le 3 and Athina Markopoulou 2 Abstract Virtual reality (VR) platforms and apps collect user sensor data, including motion, facial, eye, and hand data, in abstracted form. These data may expose users to unique privacy risks without their knowledge or meaningful awareness, yet the extent of these risks remains understudied. To address this gap, we propose VR ProfiLens , a framework to study user profiling based on VR sensor data and the resulting privacy risks across consumer VR apps. To systematically study this problem, we first develop a taxonomy rooted in CCPA definition of personal information and expanded it by sensor, app, and threat contexts to identify user attributes at risk. Then, we conduct a user study in which we collect VR sensor data from four sensor groups from real users interacting with 10 popular consumer VR apps, followed by a survey. We design and apply an analysis pipeline to demonstrate the feasibility of inferring user attributes using these data. Our results demonstrate the feasibility of user attribute inference, including sensitive personal information, have a moderately high to high risk (with up to ∼ 90 % \sim 90\% F1 score) of being inferred from the abstracted sensor data. Through feature analysis, we further identify correlations among app groups and sensor groups in inferring user attributes. Our findings highlight risks to users, including privacy loss, tracking, targeted advertising, and safety threats. Finally, we discuss both design implications and regulatory recommendations to enhance transparency and better protect users’ privacy in VR. I Introduction Virtual reality (VR) provides immersive, interactive experiences for users across diverse apps, such as gaming, education and remote work [ 1 , 2 , 3 ] . As part of the broader Metaverse, which connects virtual, augmented, and mixed reality–also known as extended reality (XR) [ 4 ] , the VR market continues to grow rapidly [ 5 ] , driven by major platforms such as the Meta [ 6 ] , SteamVR [ 7 ] . Though VR offers substantial benefits [ 1 , 2 , 3 ] , its extensive data collection and immersive nature introduce unique privacy and security challenges for users. User Privacy and Security in VR. VR apps collect diverse sets of sensor data that may contain users’ biometrics or behavioral fingerprints. Compared to mobile and web platforms, users have far less control over their privacy decisions in VR [ 8 , 9 , 10 ] . Independent reviews of major VR platforms reveal weak security controls and vague privacy policies, limiting users’ ability to make informed decisions regarding sharing their sensitive data [ 11 , 12 ] . Meta, for example, discloses in their privacy policy that platforms and app developers can collect and use “abstracted” data derived from raw inputs [ 13 ] , which may seem less privacy-sensitive. However, prior works show that sensor data 1 1 1 “Abstracted” sensor data refers to processed telemetry derived from raw sensor inputs (e.g., image, video), available to platforms and app developers [ 14 , 13 ] . We use abstracted sensor data and sensor data interchangeably. can uniquely identify users [ 15 , 16 , 17 ] . As the Metaverse incorporates advertising and marketing campaigns [ 18 ] , sensor data can be repurposed for behavioral targeting. Additionally, AI agents capable of impersonating users or influencing user judgment and behavior [ 19 ] may further expand the attack surface for manipulation and identity misuse. These privacy concerns have escalated into lawsuits and enforcement actions targeting biometric data practices in immersive ecosystems, prompting stronger regulatory scrutiny and compliance expectations [ 20 , 21 , 22 , 23 ] . Problem Statement. Due to pervasive data collection and users’ limited control over data sharing in VR, “abstracted” sensor data can be exploited for user profiling and other non-functional purposes. We define user profiling as inference of private user attributes [ 14 ] , and we use the two terms interchangeably. Understanding the extent to which users can be profiled from ‘‘abstracted’’ sensor data across consumer VR apps 2 2 2 Consumer VR apps are designed for naive purposes such as social interaction (e.g., a_{1} ) or gaming (e.g., a_{3} ), not for intentional attacks. , and the risks such profiling entails remains a critical yet understudied privacy problem.We aim to answer the following question, which is later expanded into six research questions in Section V : to what extent can VR users be profiled using only “abstracted” sensor data across consumer VR apps, and how does this impact users’ VR privacy and safety? Research Gaps. Recent studies show that “abstracted” sensor data enables unique identification [ 17 , 16 , 15 ] and user tracking across apps [ 17 , 24 ] . However, user profiling using sensor data remains understudied, and existing studies exhibit several limitations. First, while prior work has proposed a VR user attribute taxonomy [ 14 ] , their focus was VR literature only, which limits the taxonomy’s scope. As a result, this taxonomy does not explain the privacy nor regulatory significance of user attributes, nor does it capture attributes relevant to broader threat scenarios such as targeted advertising, safety and harm, or app contexts. Rather than focusing on available VR data as a basis (i.e., bottom-up approach), we construct our taxonomy from privacy law and expand it with threat scenarios and app specific attributes. We subsequently analyze which attributes are currently applicable in the VR context (i.e., top-down approach), considering that other attributes included in the taxonomy may become relevant as VR ecosystem evolve. Second, prior works focused on profiling using a single sensor (i.e., body motion [ 25 , 26 ] ), leaving the risks associated with other sensors and multi-sensor combinations underexplored. Existing studies rely on custom adversarial app [ 27 , 25 ] or a single consumer app (e.g., Beat Saber [ 28 ] [ 26 ] ), limiting diversity of user activities. Moreover, prior work underexplores the privacy and safety implications of inferred attributes and offers limited user-centric, sensor- or app-specific mitigation and regulatory guidance. Approach. To address prior research gaps, we develop VR ProfiLens , a framework for systematically investigating user profiling risks in consumer VR apps, as depicted in Figure 1 . Overall, we make the following key contributions: (1) VR User Profiling Taxonomy (Section III-D ). We introduce a novel VR User Profiling Taxonomy that is rooted in privacy law, namely the California Consumer Privacy Act’s (CCPA) definition of personal information [ 29 ] , enabling systematic reasoning about privacy and regulatory relevance of user attributes. The taxonomy is further expanded across diverse threat scenarios, including targeted advertising, identity theft, and safety and harm, as well as prior literature related to VR profiling and app contexts, enabling us to identify and analyze user attributes across different threat scenarios and app groups. Our taxonomy enables us to analyze the relationships among user attributes, sensor groups, app

Read the complete text