提出防御多模态模型隐性联合攻击的新方法，提升真实场景安全性。
CrossGuard: Safeguarding MLLMs against Joint-Modal Implicit Malicious Attacks
- 用强化学习自动生成跨领域隐性攻击样本，构建高质量数据集。
- 在14个领域测试中，防御效果显著优于现有方案，兼顾安全与可用性。
- 适合关注多模态模型安全的开发者和研究者使用。
多模态大语言模型（MLLMs）具备强大的推理与感知能力，但日益面临越狱攻击威胁。现有研究多聚焦于单模态显式攻击，而最新研究表明，良性文本与图像联合可隐含恶意意图，此类联合模态威胁难以检测且研究不足，主因是高质量隐性数据稀缺。我们提出ImpForge，一个基于强化学习与定制奖励模块的自动化红队管道，可生成覆盖14个领域的多样化隐性样本。基于该数据集，进一步开发了CrossGuard，一种意图感知的防护机制，能有效抵御显式与隐性双重威胁。在安全与非安全基准、显式与隐性攻击及多种域外设置下的广泛实验表明，CrossGuard显著优于现有防御方案，包括先进MLLMs与护栏系统，在保持高实用性的同时实现更强安全性。该工作为提升真实世界多模态模型鲁棒性提供了平衡且实用的解决方案。代码已开源：https://github.com/ZhangXu0963/CrossGuard。
原文摘要 · Abstract (English)
Multimodal Large Language Models (MLLMs) achieve strong reasoning and perception capabilities but are increasingly vulnerable to jailbreak attacks. While existing work focuses on explicit attacks, where malicious content resides in a single modality, recent studies reveal implicit attacks, in which benign text and image inputs jointly express unsafe intent. Such joint-modal threats are difficult to detect and remain underexplored, largely due to the scarcity of high-quality implicit data. We propose ImpForge, an automated red-teaming pipeline that leverages reinforcement learning with tailored reward modules to generate diverse implicit samples across 14 domains. Building on this dataset, we further develop CrossGuard, an intent-aware safeguard providing robust and comprehensive defense against both explicit and implicit threats. Extensive experiments across safe and unsafe benchmarks, implicit and explicit attacks, and multiple out-of-domain settings demonstrate that CrossGuard significantly outperforms existing defenses, including advanced MLLMs and guardrails, achieving stronger security while maintaining high utility. This offers a balanced and practical solution for enhancing MLLM robustness against real-world multimodal threats. Our code is released: https://github.com/ZhangXu0963/CrossGuard.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。