Source note2 min read
How Agents Ask for Permission: User Permissions for AI Agents, from Interfaces to Enforcement
Source
arXiv
arxiv.org/abs/2607.13718v1
Source
- Source name: arXiv
- Source kind: paper
- Authors / organization: Alexandra E. Michael, Franziska Roesner
- Venue: arXiv: cs.CR
- Published: 2026-07-15
- Collected: 2026-07-16
- URL: http://arxiv.org/abs/2607.13718v1
- PDF: https://arxiv.org/pdf/2607.13718v1
- DOI: UNKNOWN
Plain-Language Brief
The authors of this research paper review 21 proposals for agent permissions systems in AI, focusing on user-level permissions policies. They also analyze five commercial agents and compare their permissions handling to the literature. The study identifies themes across the literature and commercial agents and highlights areas for future work.
What the Document Says
- The paper discusses the increasing risks associated with AI agents and the need for user-level permissions in agentic AI systems.
- A review of 21 proposals for agent permissions systems is conducted, focusing on how they specify user-level policies, derive internal policies from user input, and enforce those policies at run-time.
- Five prominent commercial agents are analyzed for their permissions handling and compared to agentic permissions systems in the literature.
- The study identifies high-level themes across the literature and commerical agents, as well as gaps where future work is needed.
Key Concepts Mentioned
- AI agents
- user-level permissions
- agentic security
- prompt injection attacks
- hallucination
Claims Made by the Document
- Different users have different needs and preferences, necessitating support for user-level permissions policies in agentic AI systems.
- 21 proposals for agent permissions systems are reviewed in the paper.
- Five commercial agents are analyzed for their permissions handling.
Evidence Used
- The authors discuss the increasing risks associated with AI agents and the need for user-level permissions policies in agentic AI systems (in the introduction).
- The review of proposals and analysis of commercial agents is detailed throughout the paper.
Methods / Approach
UNKNOWN
Limitations Stated by the Authors
UNKNOWN
Limitations Noted by the Agent
Agent interpretation — not from the source.
UNKNOWN
Open Questions
UNKNOWN
Notes for Future Q&A
May include interpretation. Labeled where it does.
UNKNOWN
Raw Abstract / Excerpt
As AI agents gain prevalance, users are increasingly exposed to the risks such systems entail. Prompt injection attacks, as well as hallucination, can cause agents to leak private information to third parties. As autonomous systems, agents also present the more active danger of performing sensitive tasks, such as bank transactions, without the user's intent or authorization. Recognizing this challenge, the agentic security community has developed numerous proposals for secure agentic systems. Much of this work has focused on product-level approaches, where agentic system developers determine and apply the same security policies and permissions to all users. Yet different users have different needs and preferences, necessitating support for user-level permissions policies in agentic AI systems. To understand how user-level permissions are handled in AI agent systems, we survey 21 proposals for agent permissions systems. From this review, we construct a taxonomy of how different systems specify user-level permissions policies, both at the user interface and internally; derive internal policies from user input; and enforce those policies at run-time. We then analyze five prominent commercial agents and compare their permissions handling to agentic permissions systems in the literature. We identify several high-level themes across the literature and commerical agents, as well as multiple gaps where future work is needed.
