mlsanityData to 7 Oct 2026

Assessing Privacy Preservation and Utility in Online Vision-Language Models

arXiv 2604.09695 ↗First posted 6 Apr 2026cs.CV cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

The increasing use of Online Vision Language Models (OVLMs) for processing images has introduced significant privacy risks, as individuals frequently upload images for various utilities, unaware of the potential for privacy violations. Images contain relationships that relate to Personally Identifiable Information (PII), where even seemingly harmless details can indirectly reveal sensitive information through surrounding clues. This paper explores the critical issue of PII disclosure in images uploaded to OVLMs and its implications for user privacy. We investigate how the extraction of contextual relationships from images can lead to direct (explicit) or indirect (implicit) exposure of PII, significantly compromising personal privacy. Furthermore, we propose methods to protect privacy while preserving the intended utility of the images in Vision Language Model (VLM)-based applications. Our evaluation demonstrates the efficacy of these techniques, highlighting the delicate balance between maintaining utility and protecting privacy in online image processing environments. Index Terms-Personally Identifiable Information (PII), Privacy, Utility, privacy concerns, sensitive information

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 8 of its 13 references that are in the map. First posted on arXiv 6 Apr 2026.

Builds on

8 of 13 references.

References · 8

  1. 2020Extracting Training Data from Large Language Modelscited by 2,144
  2. 2022Quantifying Memorization Across Neural Language Modelscited by 755
  3. 2024WildVision: Evaluating Vision-Language Models in the Wild with Human Preferencescited by 76
  4. 2024PAPILLON: Privacy Preservation from Internet-based and Local Language Model Ensemblescited by 54
  5. 2024Private Attribute Inference from Images with Vision-Language Modelscited by 35
  6. 2020PrivacyNet: Semi-Adversarial Networks for Multi-attribute Face Privacycited by 30
  7. 2021DP-Image: Differential Privacy for Image Data in Feature Spacecited by 23
  8. 2025Generative Visual Commonsense Answering and Explaining with Generative Scene Graph Constructingcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026PII-VisBench: Evaluating Personally Identifiable Information Safety in Vision Language Models Along a Continuum of Visibility
  2. 2024The Phantom Menace: Unmasking Privacy Leakages in Vision-Language Models
  3. 2024Multi-PA: A Multi-perspective Benchmark on Privacy Assessment for Large Vision-Language Models
  4. 2025Assessing Visual Privacy Risks in Multimodal AI: A Novel Taxonomy-Grounded Evaluation of Vision-Language Models
  5. 2026Vision Language Model Helps Private Information De-Identification in Vision Data
  6. 2025Through Their Eyes: User Perceptions on Sensitive Attribute Inference of Social Media Videos by Visual Language Models
  7. 2025Defeating Cerberus: Concept-Guided Privacy-Leakage Mitigation in Multimodal Language Models
  8. 2026PrivLEX: Detecting legal concepts in images through Vision-Language Models