arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.09091cs.CRcs.AIcs.HC
收藏

与LLM玩猜谜游戏:利用隐式线索的间接越狱攻击

Play Guessing Game with LLM: Indirect Jailbreak Attack with Implicit Clues

ISCAS(中国科学院软件研究所);NTU(南洋理工大学)

Zhiyuan Chang, Mingyang Li, Yi Liu, Junjie Wang, Qing Wang, Yang Liu

AI总结:

提出间接越狱攻击方法Puzzler,通过隐式线索绕过LLM防御,在闭源LLM上查询成功率达96.6%,并有效规避检测。

摘要:

随着大语言模型(LLM)的发展,其安全威胁日益受到关注。为评估LLM的安全防御能力,研究者提出了众多越狱攻击方法。现有越狱攻击主要采用场景伪装技术,然而其明确提及的恶意意图容易被LLM识别并防御。本文提出一种间接越狱攻击方法Puzzler,该方法通过向LLM隐式提供原始恶意查询的相关线索,能够绕过LLM的防御策略并获取恶意响应。此外,受《孙子兵法》中“不可攻则守”智慧的启发,我们采取防御姿态,通过LLM收集原始恶意查询的线索。大量实验结果表明,Puzzler在闭源LLM上实现了96.6%的查询成功率,比基线方法高出57.9%-82.7%。进一步,在与最先进的越狱检测方法对抗测试中,Puzzler相比基线方法展现出更强的规避检测能力。

英文摘要:

With the development of LLMs, the security threats of LLMs are getting more and more attention. Numerous jailbreak attacks have been proposed to assess the security defense of LLMs. Current jailbreak attacks primarily utilize scenario camouflage techniques. However their explicitly mention of malicious intent will be easily recognized and defended by LLMs. In this paper, we propose an indirect jailbreak attack approach, Puzzler, which can bypass the LLM's defense strategy and obtain malicious response by implicitly providing LLMs with some clues about the original malicious query. In addition, inspired by the wisdom of "When unable to attack, defend" from Sun Tzu's Art of War, we adopt a defensive stance to gather clues about the original malicious query through LLMs. Extensive experimental results show that Puzzler achieves a query success rate of 96.6% on closed-source LLMs, which is 57.9%-82.7% higher than baselines. Furthermore, when tested against the state-of-the-art jailbreak detection approaches, Puzzler proves to be more effective at evading detection compared to baselines.

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。

↑