mlsanityData to 7 Oct 2026

Jailbreaking Multimodal Large Language Models using Multi-Clip Video

arXiv 2606.02111 ↗First posted 1 Jun 2026cs.CV cs.AI cs.CLcited by 0

On the map: Multimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness

also in 1 more
Abstract (arXiv)

As multimodal large language models (MLLMs) have advanced to process video inputs, concerns have emerged about their potential for malicious misuse. Prior jailbreak studies have shown that safety alignment in MLLMs can be bypassed through visual inputs, yet it remains unclear which properties of video inputs induce this vulnerability. To address this gap, we introduce Multi-Clip Video (MCV) SafetyBench, a dataset of 2,920 videos designed to evaluate how the diversity of video inputs affects the vulnerability of MLLMs. Each video consists of multiple short clips depicting diverse contexts related to a harmful query. Experiments on eight representative video MLLMs show that attack success consistently increases with the number of clips. Our results further indicate that the video modality is (1) more vulnerable than the image modality, (2) more vulnerable to dynamic videos than to static videos, and (3) more vulnerable when videos contain more diverse contexts. Building on these findings, we propose a defense strategy that leverages the relative robustness of the image modality.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in MLLM jailbreaks and adversarial robustness, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 38 of its 57 references that are in the map. First posted on arXiv 1 Jun 2026.

Builds on

38 of 57 references.

Closest ancestors · 9

Same direction, newest first

  1. 2025JailBound: Jailbreaking Internal Safety Boundaries of Vision-Language Models26 May
  2. 2025Video-SafetyBench: A Benchmark for Safety Evaluation of Video LVLMs17 May
  3. 2025Distraction is All You Need for Multimodal Large Language Model Jailbreaking15 Feb
  4. 2025Jailbreaking Multimodal Large Language Models via Shuffle Inconsistency9 Jan
  5. 2024VLSBench: Unveiling Visual Leakage in Multimodal Safety29 Nov
  6. 2024Llama Guard 3 Vision: Safeguarding Human-AI Image Understanding Conversations15 Nov
  7. 2024JailBreakV: A Benchmark for Assessing the Robustness of MultiModal Large Language Models against Jailbreak Attacks3 Apr
  8. 2024AdaShield: Safeguarding Multimodal Large Language Models from Structure-based Attack via Adaptive Shield Prompting14 Mar
  9. 2023FigStep: Jailbreaking Large Vision-Language Models via Typographic Visual Prompts9 Nov

Other references · 29

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2023Improved Baselines with Visual Instruction Tuningcited by 4,362
  3. 2025Wan: Open and Advanced Large-Scale Video Generative Modelscited by 2,536
  4. 2023Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyondcited by 2,179
  5. 2023Jailbroken: How Does LLM Safety Training Fail?cited by 1,565
  6. 2023Video-LLaVA: Learning United Visual Representation by Alignment Before Projectioncited by 1,385
  7. 2024Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysiscited by 1,375
  8. 2023Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!cited by 1,106
  9. 2023Shikra: Unleashing Multimodal LLM's Referential Dialogue Magiccited by 891
  10. 2023Vision-Language Models for Vision Tasks: A Surveycited by 834
  11. 2024LLaVA-Video: Video Instruction Tuning With Synthetic Datacited by 550
  12. 2024How Johnny Can Persuade LLMs to Jailbreak Them: Rethinking Persuasion to Challenge AI Safety by Humanizing LLMscited by 533
17 more
  1. 2023PandaGPT: One Model To Instruction-Follow Them Allcited by 330
  2. 2023Shadow Alignment: The Ease of Subverting Safely-Aligned Language Modelscited by 284
  3. 2022Revisiting the "Video" in Video-Language Understandingcited by 167
  4. 2024Streaming Long Video Understanding with Large Language Modelscited by 164
  5. 2023BLIVA: A Simple Multimodal LLM for Better Handling of Text-Rich Visual Questionscited by 125
  6. 2024HourVideo: 1-Hour Video-Language Understandingcited by 110
  7. 2023How Many Unicorns Are in This Image? A Safety Evaluation Benchmark for Vision LLMscited by 103
  8. 2024Natural Language Understanding and Inference with MLLM in Visual Question Answering: A Surveycited by 74
  9. 2019Only Time Can Tell: Discovering Temporal Data for Temporal Modelingcited by 66
  10. 2024Jailbreak Large Vision-Language Models Through Multi-Modal Linkagecited by 53
  11. 2024Jailbreak Attacks and Defenses against Multimodal Generative Models: A Surveycited by 31
  12. 2024Personalizing Multimodal Large Language Models for Image Captioning: An Experimental Analysiscited by 23
  13. 2024ActionAtlas: A VideoQA Benchmark for Domain-specialized Action Recognitioncited by 21
  14. 2025The Devil is in the Prompts: Retrieval-Augmented Prompt Optimization for Text-to-Video Generationcited by 21
  15. 2025FC-Attack: Jailbreaking Multimodal Large Language Models via Auto-Generated Flowchartscited by 15
  16. 2023A Large-scale Study of Spatiotemporal Representation Learning with a New Benchmark on Action Recognitioncited by 14
  17. 2024FMM-Attack: A Flow-based Multi-modal Adversarial Attack on Video-based LLMscited by 14

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026SceneJail: Exploiting Video Scenario Context to Jailbreak Multimodal LLMs
  2. 2024Images are Achilles' Heel of Alignment: Exploiting Visual Vulnerabilities for Jailbreaking Multimodal Large Language Models
  3. 2024MVTamperBench: Evaluating Robustness of Vision-Language Models
  4. 2025Do We Really Need Curated Malicious Data for Safety Alignment in Multi-modal Large Language Models?
  5. 2025Align is not Enough: Multimodal Universal Jailbreak Attack against Multimodal Large Language Models
  6. 2025PolyJailbreak: Cross-Modal Jailbreaking Attacks on Black-Box Multimodal LLMs
  7. 2025From Evaluation to Defense: Advancing Safety in Video Large Language Models
  8. 2023MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large Language Models