跨模态混淆：针对大型视觉-语言模型的越狱攻击
Cross-Modal Obfuscation for Jailbreak Attacks on Large Vision-Language Models
AI总结：
针对大型视觉-语言模型，提出跨模态混淆攻击框架CAMO，通过分解恶意提示为良性片段并利用跨模态推理重构指令，实现高效隐蔽的越狱攻击，揭示现有安全机制漏洞。
摘要：
大型视觉-语言模型（LVLMs）在多模态任务中展现出卓越性能，但仍易受越狱攻击影响，此类攻击绕过内置安全机制以诱导生成受限内容。现有的黑盒越狱方法主要依赖对抗性文本提示或图像扰动，然而这些方法极易被标准内容过滤系统检测，且查询和计算效率较低。在本工作中，我们提出跨模态对抗性多模态混淆（CAMO），一种新颖的黑盒越狱攻击框架，将恶意提示分解为语义良性的视觉和文本片段。通过利用LVLMs的跨模态推理能力，CAMO通过多步推理隐蔽地重构有害指令，从而规避常规检测机制。我们的方法支持可调节的推理复杂度，且相比先前攻击所需查询次数显著减少，实现了隐蔽性与效率兼得。在领先的LVLMs上进行的全面评估验证了CAMO的有效性，展示了稳健的性能和强跨模型迁移性。这些结果凸显了当前内置安全机制中的重大漏洞，强调了在视觉-语言系统中迫切需要先进的对齐感知安全与保障解决方案。
英文摘要：
Large Vision-Language Models (LVLMs) demonstrate exceptional performance across multimodal tasks, yet remain vulnerable to jailbreak attacks that bypass built-in safety mechanisms to elicit restricted content generation. Existing black-box jailbreak methods primarily rely on adversarial textual prompts or image perturbations, yet these approaches are highly detectable by standard content filtering systems and exhibit low query and computational efficiency. In this work, we present Cross-modal Adversarial Multimodal Obfuscation (CAMO), a novel black-box jailbreak attack framework that decomposes malicious prompts into semantically benign visual and textual fragments. By leveraging LVLMs' cross-modal reasoning abilities, CAMO covertly reconstructs harmful instructions through multi-step reasoning, evading conventional detection mechanisms. Our approach supports adjustable reasoning complexity and requires significantly fewer queries than prior attacks, enabling both stealth and efficiency. Comprehensive evaluations conducted on leading LVLMs validate CAMO's effectiveness, showcasing robust performance and strong cross-model transferability. These results underscore significant vulnerabilities in current built-in safety mechanisms, emphasizing an urgent need for advanced, alignment-aware security and safety solutions in vision-language systems.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
