English(EN) Benchmarking Web Agent Safety under E-commerce Deceptive Interfaces

新基准测试探究AI代理在欺骗性界面和不安全操作下的安全性

作者 PulseAugur 编辑部 · [3 个来源] · 2026-06-15 04:00

两篇新的研究论文介绍了用于评估AI代理安全性的基准测试。OSGuard专注于计算机使用代理，区分安全和不安全的操作，并识别任务执行中的潜在危险。WebDecept针对网络代理，专门测试它们在电子商务场景中对欺骗性界面的易感性，发现当前代理存在漏洞，基于提示的约束通常不足。 AI

影响这些基准测试突显了当前AI代理在安全性方面的关键差距，尤其是在欺骗性界面和不安全捷径方面，敦促在稳健的实际部署方面进行进一步研究。

排序理由两篇在arXiv上发表的学术论文，介绍了用于AI代理安全性的新基准测试。

AI 生成摘要 · Google Gemini · 来自 3 个来源。我们如何撰写摘要 →

报道来源 [3]

arXiv cs.AI TIER_1 English(EN) · Guruprasad Viswanathan Ramesh, Asmit Nayak, Basieem Siddique, Kassem Fawaz · 2026-06-18 04:00

WebSP-Eval：评估网站安全和隐私任务上的Web代理

arXiv:2604.06367v2 Announce Type: replace-cross Abstract: Web agents automate browser tasks, ranging from simple form completion to complex workflows like ordering groceries. While current benchmarks evaluate general-purpose performance~(e.g., WebArena) or safety against maliciou…
arXiv cs.AI TIER_1 English(EN) · Mina Mohammadmirzaei, Jeffrey Flanigan · 2026-06-16 04:00

OSGuard：计算机使用代理的安全基准

arXiv:2606.15034v1 Announce Type: new Abstract: Computer-use agents are increasingly evaluated by whether they complete realistic desktop and web tasks. However, task success alone can miss failures in which an agent reaches the nominal goal through an unsafe shortcut. We introdu…
arXiv cs.CL TIER_1 English(EN) · Zijing Shi, Meng Fang, Ling Chen · 2026-06-15 04:00

电商欺骗性界面下的网络代理安全基准测试

arXiv:2606.13686v1 Announce Type: new Abstract: As autonomous web agents are increasingly deployed to perform real-world tasks, ensuring their safety has become a critical concern. In this work, we study web agent behavior under realistic deceptive interfaces in the e-commerce do…