English(EN) LLMail-Inject: A Dataset from a Realistic Adaptive Prompt Injection Challenge

新的LLMail-Inject数据集揭示了自适应提示注入攻击的高成功率

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-12 04:00

一个名为LLMail-Inject的新数据集已被发布，其中包含超过20万个由人类编写的提示，旨在测试大型语言模型上的自适应提示注入攻击。该数据集旨在通过捕捉真实世界对手的复杂性和适应性来解决现有基准的局限性。初步评估显示，即使是专门针对提示注入进行训练的模型（如Meta SecAlign）以及像GPT-4o和GPT-5 mini这样的先进模型，在执行来自不可信来源的未经授权任务方面也表现出很高的成功率。 AI

影响突出了LLM的关键漏洞，可能推动新的安全研究和防御策略。

排序理由该集群描述了一篇介绍用于评估LLM针对提示注入攻击安全性的数据集的新学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 11 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群描述了一篇介绍用于评估LLM针对提示注入攻击安全性的数据集的新学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, safety; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Sahar Abdelnabi, Aideen Fay, Ahmed Salem, Egor Zverev, Kai-Chieh Liao, Chi-Huang Liu, Chun-Chih Kuo, Jannis Weigend, Danyael Manlangit, Alex Apostolov, Haris Umair, Jo\~ao Donato, Masayuki Kawakita, Athar Mahboob, Tran Huu Bach, Tsun-Han Chiang, Myeongji… · 2026-10-12 04:00

LLMail-Inject：来自真实自适应提示注入挑战的数据集

arXiv:2506.09956v2 Announce Type: replace-cross Abstract: Indirect prompt injection attacks exploit a fundamental weakness of LLMs: executing unauthorized tasks from untrusted sources. Despite the practical significance of this attack vector, existing benchmarks rely on small set…

本文如何被排名

报道来源 [1]

LLMail-Inject：来自真实自适应提示注入挑战的数据集

相关实体

相关话题