# VisualLeakBench: Auditing the Fragility of Large Vision-Language Models against PII Leakage and Social Engineering

arXiv: https://arxiv.org/abs/2603.13385 · HTML: https://mlsanity.com/paper/2603.13385/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in MLLM jailbreaks and adversarial robustness, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 17 of its 20 references that are in the map.
- First posted on arXiv 11 Mar 2026.

## Placed in

- Multimodal LLMs / vision-language models / [MLLM jailbreaks and adversarial robustness](https://mlsanity.com/mllm/mllm-jailbreaks-and-adversarial-robustness/)
- LLM safety and security / [Vision-language and text-to-image safety](https://mlsanity.com/safety/vision-language-and-text-to-image-safety/)

## Builds on (most cited first, 17 of 17)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Universal and Transferable Adversarial Attacks on Aligned Language Models](https://mlsanity.com/paper/2307.15043/) (2023)
- [Constitutional AI: Harmlessness from AI Feedback](https://mlsanity.com/paper/2212.08073/) (2022)
- [Jailbroken: How Does LLM Safety Training Fail?](https://mlsanity.com/paper/2307.02483/) (2023)
- [Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection](https://mlsanity.com/paper/2302.12173/) (2023)
- [Ignore Previous Prompt: Attack Techniques For Language Models](https://mlsanity.com/paper/2211.09527/) (2022)
- [Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned](https://mlsanity.com/paper/2209.07858/) (2022)
- [Formalizing and Benchmarking Prompt Injection Attacks and Defenses](https://mlsanity.com/paper/2310.12815/) (2023)
- [Visual Adversarial Examples Jailbreak Aligned Large Language Models](https://mlsanity.com/paper/2306.13213/) (2023)
- [R-Judge: Benchmarking Safety Risk Awareness for LLM Agents](https://mlsanity.com/paper/2401.10019/) (2024)
- [Analyzing Leakage of Personally Identifiable Information in Language Models](https://mlsanity.com/paper/2302.00539/) (2023)
- [MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large Language Models](https://mlsanity.com/paper/2311.17600/) (2023)
- [Adversarial GLUE: A Multi-Task Benchmark for Robustness Evaluation of Language Models](https://mlsanity.com/paper/2111.02840/) (2021)
- [JailBreakV: A Benchmark for Assessing the Robustness of MultiModal Large Language Models against Jailbreak Attacks](https://mlsanity.com/paper/2404.03027/) (2024)
- [Image Hijacks: Adversarial Images can Control Generative Models at Runtime](https://mlsanity.com/paper/2309.00236/) (2023)
- [A Survey of Safety on Large Vision-Language Models: Attacks, Defenses and Evaluations](https://mlsanity.com/paper/2502.14881/) (2025)
- [PrivLM-Bench: A Multi-level Privacy Evaluation Benchmark for Language Models](https://mlsanity.com/paper/2311.04044/) (2023)

## Built on by (most cited first, 4 of 4)

- [Sustainable Hybrid Document-Routed Retrieval for Financial RAG: Resolving the Robustness-Precision Trade-off](https://mlsanity.com/paper/2603.26815/) (2026)
- [MaskClaw: Edge-Side Personalized Privacy Arbitration for GUI Agents with Behavior-Driven Skill Evolution](https://mlsanity.com/paper/2605.28646/) (2026)
- [Energy-Efficient On-Device RAG on a Mobile NPU: System Design and Benchmark on Snapdragon X Elite](https://mlsanity.com/paper/2606.11257/) (2026)
- [VisualLeakBench: Reproducible Action-Boundary Propagation Failures in Vision-Language Agents](https://mlsanity.com/paper/2606.07595/) (2026)

## Abstract (arXiv)

As Large Vision-Language Models (LVLMs) are increasingly deployed in agent-integrated workflows and other deployment-relevant settings, their robustness against semantic visual attacks remains under-evaluated -- alignment is typically tested on explicit harmful content rather than privacy-critical multimodal scenarios. We introduce VisualLeakBench, an evaluation suite to audit LVLMs against OCR Injection and Contextual PII Leakage using 1,000 synthetically generated adversarial images with 8 PII types, validated on 50 in-the-wild (IRL) real-world screenshots spanning diverse visual contexts. We evaluate four frontier systems (GPT-5.2, Claude~4, Gemini-3 Flash, Grok-4) with Wilson 95% confidence intervals. Claude~4 achieves the lowest OCR ASR (14.2%) but the highest PII ASR (74.4%), exhibiting a comply-then-warn pattern -- where verbatim data disclosure precedes any safety-oriented language. Grok-4 achieves the lowest PII ASR (20.4%). A defensive system prompt eliminates PII leakage for two models, reduces Claude~4's leakage from 74.4% to 2.2%, but has no effect on Gemini-3 Flash on synthetic data. Strikingly, IRL validation reveals Gemini-3 Flash does respond to mitigation on real-world images (50% to 0%), indicating that mitigation robustness is template-sensitive rather than uniformly absent. We release our dataset and code for reproducible robustness and safety evaluation of deployment-relevant vision-language systems.