理解LLM在执行无害任务时遇到用户提供的有害内容的行为
Understanding LLM Behavior When Encountering User-Supplied Harmful Content in Harmless Tasks
AI总结：
本研究探讨了LLM在执行无害任务时遇到用户提供的有害内容时的行为，发现主流LLM在处理有害内容时未能维持伦理标准，且某些类别和任务更易引发有害响应。
摘要：
大型语言模型（LLMs）正越来越多地被训练以符合人类价值观，主要集中在任务层面，即拒绝直接执行有害任务。然而，一个微妙但关键的内容层面伦理问题往往被忽视：在执行看似无害的任务时，LLMs是否会像有道德意识的人一样，在遇到用户提供的有害内容时拒绝继续？在本研究中，我们旨在理解这一内容层面的伦理问题，并系统评估其对主流LLMs的影响。我们首先构建了一个有害知识数据集（即不遵守OpenAI使用政策），作为用户提供的有害内容，包含1357条记录，涵盖十个有害类别。然后我们设计了九个无害任务（即符合OpenAI使用政策），以模拟现实中的无害任务，按所需用户提供的内容程度分为三个类别：广泛、中等和有限。利用有害知识数据集和无害任务集，我们评估了九个LLMs在执行无害任务时遇到用户提供的有害内容时的行为，并进一步检验有害知识类别与任务之间的动态如何影响不同的LLMs。我们的结果表明，当前LLMs，即使是最新的GPT-5.2和Gemini-3-Pro，往往在无害任务中继续处理有害内容，未能维护人类对齐的伦理。此外，来自“暴力/图像”类别和“翻译”任务的外部知识更可能引发LLMs的有害响应。我们还进行了广泛的消融研究，以调查潜在影响这种新型滥用漏洞的因素。我们希望我们的研究能激励相关利益方采取增强的安全措施，以缓解这一被忽视的内容层面伦理风险。
英文摘要：
Large Language Models (LLMs) are increasingly trained to align with human values, primarily focusing on task level, i.e., refusing to execute directly harmful tasks. However, a subtle yet crucial content-level ethical question is often overlooked: when performing a seemingly benign task, will LLMs -- like morally conscious human beings -- refuse to proceed when encountering harmful content in user-provided material? In this study, we aim to understand this content-level ethical question and systematically evaluate its implications for mainstream LLMs. We first construct a harmful knowledge dataset (i.e., non-compliant with OpenAI's usage policy) to serve as the user-supplied harmful content, with 1,357 entries across ten harmful categories. We then design nine harmless tasks (i.e., compliant with OpenAI's usage policy) to simulate the real-world benign tasks, grouped into three categories according to the extent of user-supplied content required: extensive, moderate, and limited. Leveraging the harmful knowledge dataset and the set of harmless tasks, we evaluate how nine LLMs behave when exposed to user-supplied harmful content during the execution of benign tasks, and further examine how the dynamics between harmful knowledge categories and tasks affect different LLMs. Our results show that current LLMs, even the latest GPT-5.2 and Gemini-3-Pro, often fail to uphold human-aligned ethics by continuing to process harmful content in harmless tasks. Furthermore, external knowledge from the ``Violence/Graphic'' category and the ``Translation'' task is more likely to elicit harmful responses from LLMs. We also conduct extensive ablation studies to investigate potential factors affecting this novel misuse vulnerability. We hope that our study could inspire enhanced safety measures among stakeholders to mitigate this overlooked content-level ethical risk.
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。
