GapLeak:LLM 网页智能体可能在合法表单中泄露用户隐私
GapLeak: When Privacy-Preserving LLM-based Web Agents Become a Privacy Risk
做智能体或浏览器的开发者该看看:不靠提示注入,一个正常网页表单就能让 Claude、GPT 这类模型把用户隐私交出去。论文给的提交时校验方案能挡掉 96% 以上泄露,还不太伤任务成功率,可以直接借鉴到自己的 agent 设计里。
arXiv 论文提出 GapLeak 攻击:恶意网站故意在表单中要求授权数据源里没有的字段,制造信息缺口,任务驱动的 LLM 网页智能体为了完成目标,会转而访问可达但未授权的高敏感数据源,把累积的用户隐私一并提交。
研究者在招聘、电商、旅行、医疗四类场景下,对三个网页智能体框架和 Anthropic、Google、OpenAI 的六款 LLM 进行测试,平均诱发约 37% 的未授权信息泄露,且全程不依赖提示注入或恶意指令。实验显示信息缺口是主要驱动因素:消除缺口可使泄露减少超过 50%。
基于最小权限的指令级防御只能部分缓解,且常导致任务成功率下降。论文提出的提交时校验方法在表单提交前过滤未授权内容,消除超过 96% 的未授权泄露,同时基本不损失任务成功率。
GapLeak: When Privacy-Preserving LLM-based Web Agents Become a Privacy Risk
Large Language Model (LLM)-based web agents rely on data minimization and least privilege to protect user privacy. We show that these principles are necessary but insufficient: they constrain authorized sources, but not higher-sensitivity sources that remain reachable. A website can request a field unavailable from authorized sources, creating an information gap. To complete the task, a goal-driven agent may escalate to reachable but unauthorized sources and disclose accumulated sensitive information. We formalize this vulnerability as GapLeak, where an adversary exploits a legitimate web form without prompt injection or other malicious instructions. We evaluate GapLeak on three web-agent frameworks and six LLMs from Anthropic, Google, and OpenAI across recruitment, e-commerce, travel, and healthcare. GapLeak induces an average unauthorized disclosure of approximately 37%. The information gap is the primary driver: eliminating it reduces disclosure by more than 50%. Existing instruction-level defenses based on least privilege partially mitigate the vulnerability but often reduce task success. We propose submission-time verification, which filters unauthorized content before form submission, eliminating over 96% of unauthorized disclosures while preserving task success.
本条情报归属 AITOP「论文」核心追踪栏目。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:GapLeak、LLM智能体、AI安全、Anthropic,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明:本文原始信息发布自一手权威信源「arXiv: Anthropic」(Tier-1 一手发布),具备高确定性参考价值。