arXiv 2610.30418 · cs.CR · 2026-10-21
Agent Mosaic: Recognizing Confidentiality Leaks in Agentic LLM Systems
Daniel Ruiz, Luca Luceri, Emilio Ferrara
University of Southern California
在最新榜 #370 本站收录于 2026-10-26
一句话要点
论文量化了智能体的「马赛克泄密」：从多份可访问文档拼出敏感结论。即使是良性提问，表现最好的模型在模拟中仍有 7.08% 的回答泄露禁止的组合信息；加分级标头和聚合规则后，仍有 32.37% 的模拟回答不安全。
The paper quantifies the agent "mosaic effect" of assembling sensitive conclusions from multiple documents: even under benign queries the best model leaks forbidden composites in 7.08% of simulated answers, and with classification headers and aggregation rules 32.37% of answers remain unsafe.
阅读 arXiv 原文 PDF Hugging Face 讨论
主题与学科
同主题：智能体
- Raven: The Harness of Harnesses for Composable Agentic Intelligence ▲748
- False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents ▲642
- Omni-IO Skills: Harnessing Your Agent Omni-Native ▲485
- BizAgentBench: Unveiling Hidden Challenges in Deploying Long-Horizon Agents for Real-World Business ▲388
- Player-1: Learning to Play from Synthetic Experience ▲347
同主题：安全与可解释
- Post-Training Leaves Behavioral Shadows on Unrelated Decisions 2026-09-24 · ▲273
- S-Space: Exploring Spatial Workspace in Multimodal Language Models 2026-10-09 · ▲270
- Right Screen, Wrong Transition: World Models as Verifiers for GUI Agents 2026-10-08 · ▲188
- Behavioral Compositionality of Sparse Autoencoder Representations 2026-10-15 · ▲108
- DecepEval: A Benchmark for Evaluating Deception in LLM Agents 2026-10-06 · ▲85
同机构：University of Southern California
数据说明
元数据来自 arXiv API（已核实）；点赞来自 Hugging Face 每日论文；引用来自 Semantic Scholar。最近刷新 2026-10-26。
要点由本站根据摘要自写，不转载摘要原文，请阅读原文。