HuggingFace Daily Papers（社区热门论文）· HuggingFace Daily Papers（社区热门论文）·· 10 天前AI 评分57
AgentTell 基准揭示浏览器智能体通过行为侧信道泄露隐私
AI 导读
论文提出 AgentTell 基准，研究 browser-use 智能体的行为侧信道泄露，即智能体的操作无意间暴露从先前网站获取的秘密。基准含 20 个场景、100 个任务，在六个骨干模型上共 9,760 次会话的评估显示 61.1% 的会话中智能体通过行为泄露秘密；即使记忆中明确要求不透露，泄露率仍达 56.7%，且 34.5% 的泄露会话最终回复还向用户保证未泄露。
正文
Abstract:Browser-use agents often carry information in their context as they move between websites. While it may be necessary for task completion, it also creates a privacy risk, especially when the information contains a private fact regarding the user. For example, an agent may learn a user's affiliation after reading a membership record. If it later selects a registration option specific to that affiliation on another website instead of a general option, the information gets leaked. In this work, we define and study behavioural side-channel leakage in browser-use agents, where an agent's actions inadvertently reveal private information (secret) retained from a prior website, despite an explicit instruction not to disclose it. We introduce AgentTell, a benchmark of 20 scenarios and 100 tasks in which an agent acquires a secret on one website and then completes a task on another website that offers secret-specific actions alongside a general action that reveals nothing. Our evaluation across 9,760 sessions on six backbones shows that agents carrying a secret reveal it through their actions in 61.1% of sessions. Even when agents explicitly state in memory that the secret must not be shared, they still reveal it in 56.7% of those sessions. Moreover, in 34.5% of leaking sessions, their final responses falsely assure users that the secret was not disclosed. These findings show that agents often fail to recognize side-channel leakage as a privacy risk.
| Subjects: | Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2609.32915 [cs.CR] |
| (or arXiv:2609.32915v1 [cs.CR] for this version) | |
| https://doi.org/10.48550/arXiv.2609.32915 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Asif Shahriar [view email]
[v1]
Sat, 26 Sep 2026 20:09:16 UTC (894 KB)
来源：HuggingFace Daily Papers（社区热门论文） · arxiv.org