FrontierBreach 基准揭示企业 AI 智能体工作流的攻击成功率差异
Curiously, models from China have a higher ASR than models from the US on average, which was disproportionately due to their performance against arbitrary code
AI 导读
PromptArmor 发布 FrontierBreach,评测企业 AI 智能体工作流中的间接提示词注入攻击,九个行业的平均攻击成功率为 34.9%,医疗与社会援助工作流达到 44.4%。中国模型的平均 ASR 高于美国模型,差异主要来自任意代码执行、网络钓鱼和数据破坏任务;它们在数据窃取、勒索软件和持久化攻击中也表现出更高的攻击成功率。
正文
Curiously, models from China have a higher ASR than models from the US on average, which was disproportionately due to their performance against arbitrary code execution, phishing, and data destruction tasks.
These models performed better than their US counterparts on Data Theft, Ransomware, and Persistence attacks.
AI 热点