The Pervasive Blind Spot: Benchmarking VLM Inference Risks on Everyday Personal Videos
On the map: Multimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
Abstract (arXiv)
The proliferation of Vision-Language Models (VLMs) introduces profound privacy risks from personal videos. This paper addresses the critical yet unexplored inferential privacy threat, the risk of inferring sensitive personal attributes over the data. To address this gap, we crowdsourced a dataset of 508 everyday personal videos from 58 individuals. We then conducted a benchmark study evaluating VLM inference capabilities against human performance. Our findings reveal three critical insights: (1) VLMs possess superhuman inferential capabilities, significantly outperforming human evaluators, leveraging a shift from object recognition to behavioral inference from temporal streams. (2) Inferential risk is strongly correlated with factors such as video characteristics and prompting strategies. (3) VLM-driven explanation towards the inference is unreliable, as we revealed a disconnect between the model-generated explanations and evidential impact, identifying ubiquitous objects as misleading confounders.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in MLLM jailbreaks and adversarial robustness, one of 16 directions in Multimodal LLMs / vision-language models. First posted on arXiv 4 Nov 2025.
Builds on
No reference list yet.
Built on by
Citing papers per year: 2026: 1.
1 paper.
Similar, not cited
Close in wording, no citation either way
- 2025Through Their Eyes: User Perceptions on Sensitive Attribute Inference of Social Media Videos by Visual Language Models
- 2024Private Attribute Inference from Images with Vision-Language Models
- 2024The Phantom Menace: Unmasking Privacy Leakages in Vision-Language Models
- 2025MultiPriv: Benchmarking Individual-Level Privacy Reasoning in Vision-Language Models
- 2026How Far Are VLMs from Privacy Awareness in the Physical World? An Empirical Study
- 2026Defending Wearable VLMs Against Private Attribute Inference
- 2025Who Can See Through You? Adversarial Shielding Against VLM-Based Attribute Inference Attacks
- 2025Rethinking Post-Unlearning Behavior of Large Vision-Language Models