通过意图欺骗突破前沿基础模型
Jailbreaking Frontier Foundation Models Through Intention Deception
AI总结:
本文提出一种多轮欺骗方法,通过模拟良性意图和模型一致性,引导模型生成有害输出,并揭示了未被注意到的para-jailbreaking漏洞。
摘要:
大型(视觉-)语言模型展现出卓越能力,但极易受到突破攻击。现有安全训练方法旨在让模型学习安全与不安全之间的拒绝边界,基于用户意图。然而,这种二元训练制度常导致脆弱性,因为用户意图难以可靠评估,尤其当攻击者伪装意图时。为此,前沿模型如GPT-5已从拒绝式安全措施转向安全完成,旨在在遵守安全约束的同时最大化帮助性。然而,安全完成可能在用户伪装善意意图时被利用。具体而言,这种意图反转在多轮对话中尤为有效,因为攻击者有多次机会强化其伪装的善意意图。在本文中,我们引入了一种新颖的多轮突破方法,利用这一漏洞。我们的方法通过模拟看似良性意图并利用模型的一致性属性,逐步建立对话信任,最终引导目标模型生成有害的详细输出。最关键的是,我们的方法还揭示了一种未被注意到的模型漏洞,称为para-jailbreaking,即模型可能不会直接回复有害内容,但所揭示的信息仍然有害。我们的贡献有三方面:首先,它在包括GPT-5-thinking和Claude-Sonnet-4.5在内的前沿模型上实现了高成功率。其次,我们的方法揭示并解决了para-jailbreaking有害输出。第三,多模态VLM模型的实验表明,我们的方法优于最先进的模型。
英文摘要:
Large (vision-)language models exhibit remarkable capability but remain highly susceptible to jailbreaking. Existing safety training approaches aim to have the model learn a refusal boundary between safe and unsafe, based on the user's intent. It has been found that this binary training regime often leads to brittleness, since the user intent cannot reliably be evaluated, especially if the attacker obfuscates their intent, and also makes the system seem unhelpful. In response, frontier models, such as GPT-5, have shifted from refusal-based safeguards to safe completion, that aims to maximize helpfulness while obeying safety constraints. However, safe completion could be exploited when a user pretends their intention is benign. Specifically, this intent inversion would be effective in multi-turn conversation, where the attacker has multiple opportunities to reinforce their deceptively benign intent. In this work, we introduce a novel multi-turn jailbreaking method that exploits this vulnerability. Our approach gradually builds conversational trust by simulating benign-seeming intentions and by exploiting the consistency property of the model, ultimately guiding the target model toward harmful, detailed outputs. Most crucially, our approach also uncovered an additional class of model vulnerability that we call para-jailbreaking that has been unnoticed up to now. Para-jailbreaking describes the situation where the model may not reveal harmful direct reply to the attack query, however the information that it reveals is nevertheless harmful. Our contributions are threefold. First, it achieves high success rates against frontier models including GPT-5-thinking and Claude-Sonnet-4.5. Second, our approach revealed and addressed para-jailbreaking harmful output. Third, experiments on multimodal VLM models showed that our approach outperformed state-of-the-art models.
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。