# HarassGuard: Detecting Harassment Behaviors in Social Virtual Reality with Vision-Language Models

arXiv: https://arxiv.org/abs/2604.00592 · HTML: https://mlsanity.com/paper/2604.00592/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in MLLM jailbreaks and adversarial robustness, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 4 of its 36 references that are in the map.
- First posted on arXiv 1 Apr 2026.

## Placed in

- Multimodal LLMs / vision-language models / [MLLM jailbreaks and adversarial robustness](https://mlsanity.com/mllm/mllm-jailbreaks-and-adversarial-robustness/)

## Builds on (most cited first, 4 of 4)

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Unique Identification of 50,000+ Virtual Reality Users from Head & Hand Motion Data](https://mlsanity.com/paper/2302.08927/) (2023)
- [A Large-Scale Study of Personal Identifiability of Virtual Reality Motion Over Time](https://mlsanity.com/paper/2303.01430/) (2023)
- [Enabling Developers, Protecting Users: Investigating Harassment and Safety in VR](https://mlsanity.com/paper/2403.05499/) (2024)

## Abstract (arXiv)

Social Virtual Reality (VR) platforms provide immersive social experiences but also expose users to serious risks of online harassment. Existing safety measures are largely reactive, while proactive solutions that detect harassment behavior during an incident often depend on sensitive biometric data, raising privacy concerns. In this paper, we present HarassGuard, a vision-language model (VLM) based system that detects physical harassment in social VR using only visual input. We construct an IRB-approved harassment vision dataset, apply prompt engineering, and fine-tune VLMs to detect harassment behavior by considering contextual information in social VR. Experimental results demonstrate that HarassGuard achieves competitive performance compared to state-of-the-art baselines (i.e., LSTM/CNN, Transformer), reaching an accuracy of up to 88.09% in binary classification and 68.85% in multi-class classification. Notably, HarassGuard matches these baselines while using significantly fewer fine-tuning samples (200 vs. 1,115), offering unique advantages in contextual reasoning and privacy-preserving detection.