arXiv cs.AI· Yutao Wu, Xiao Liu, Ming Wen, Hanxun Huang, Jun Wang, Xingjun Ma·· 11 小时前AI 评分42
ScamWorld：评估 Web 智能体应对网络诈骗能力的交互式环境
ScamWorld: An Interactive Environment for Evaluating Web Agents Against Online Fraud
AI 导读
研究者推出 ScamWorld 交互式环境，重建已报告的诈骗网站以系统评估 Web 智能体。对六个前沿 LLM 驱动的智能体测试显示，基线受骗率为 11% 至 35%，且智能体更易执行付款而非提供凭证或钱包信息。轨迹分析发现部分智能体已察觉可疑情况仍继续操作，有时将用户指令视为接受网站附加条款的授权。
正文 · 原文
Abstract:Web agents increasingly browse, shop, and manage accounts on behalf of their users. Prior work has shown that they can be misled by adversarial content, including prompt injections and deceptive interfaces. Beyond such attacks, agents operating on the open web also encounter online fraud designed to deceive humans, including fraudulent online stores, fake investment platforms, and advance-fee schemes. How web agents respond to such fraud during task execution remains underexplored, in part because scam websites are short-lived, making systematic evaluation difficult. To address this gap, we introduce ScamWorld, an interactive environment that reconstructs reported scam websites across diverse scenarios, fraud mechanisms, and languages. Each reconstructed website reproduces documented fraud mechanisms, making relevant signs of fraud observable before the agent's final action and enabling controlled, repeatable evaluation. Through an extensive study of web agents powered by six frontier LLMs, we find that baseline deception rates range from 11% to 35% across models. Agents proceed more readily with payments than with requests for credentials or wallet information. A closer analysis of agent trajectories shows that some agents acknowledge suspicious conditions yet proceed, at times treating the user's instructions as authorisation to accept additional terms imposed by the website. ScamWorld provides a reusable testbed for systematically evaluating and stress-testing web agents against real-world online fraud.
| Comments: | 15 pages, 17 figures, 7 tables. Code: this https URL |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.04879 [cs.AI] |
| (or arXiv:2611.04879v2 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.04879 arXiv-issued DOI via DataCite |
Submission history
From: Oscar Wu [view email]
[v1]
Tue, 3 Nov 2026 13:31:11 UTC (14,977 KB)
[v2]
Wed, 4 Nov 2026 07:01:24 UTC (14,977 KB)
来源：arXiv cs.AI · arxiv.org