arXiv cs.AI· Jinhao Zhu, Xiao Huang, Kevin Tseng, Gil Vernik, Shishir G. Patil, Vivian Fang, Raluca Ada Popa·· 10 天前AI 评分42
MiniScope：以最小权限授权 AI 智能体
MiniScope: Authorizing Agents with Least-Privilege Permissions
AI 导读
针对 AI 智能体权限管理难题，研究者提出以任务为中心的分层权限模型并构建端到端权限系统 MiniScope，自动发现权限层级并在运行时执行上下文最小权限。评估显示，MiniScope 将谨慎型和典型用户画像的模拟权限确认减少 43.4%-89.4%，并缓解所有权限提升攻击，对实用性和运行时影响可忽略。
正文 · 原文
Abstract:AI agents are increasingly granted autonomous access to sensitive user data and third-party services, making effective permission management a critical security challenge. Existing permission models, however, typically rely on flat permission structures that fail to balance security with usability: fine-grained confirmation induces user fatigue, while coarse-grained or persistent approval leads to overprivileged agents. To address this tradeoff, we propose a task-centric, hierarchical permission model that treats an agent as a delegate operating within a task-specific role instead of requiring a separate permission decision for every tool call. Building on this model, we present MiniScope, an end-to-end permission system for agents that automates permission-hierarchy discovery and enforces contextual least privilege at runtime. Our evaluation shows that MiniScope reduces simulated permission confirmations by 43.4%-89.4% for cautious and typical personas relative to per-tool prompting and mitigates all privilege-escalation attacks with negligible impact on utility and runtime. Applied to real-world deployments, MiniScope further uncovers six overprivileged connector configurations in ChatGPT and Claude.
| Subjects: | Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2512.11147 [cs.CR] |
| (or arXiv:2512.11147v2 [cs.CR] for this version) | |
| https://doi.org/10.48550/arXiv.2512.11147 arXiv-issued DOI via DataCite |
Submission history
From: Jinhao Zhu [view email]
[v1]
Thu, 11 Dec 2025 22:10:39 UTC (2,269 KB)
[v2]
Tue, 6 Oct 2026 07:56:06 UTC (2,011 KB)
来源：arXiv cs.AI · arxiv.org