Cover Image

The Authorization-Execution Gap Is a Major Safety and Security Problem in Open-World Agents

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Wu, Baoyuan, Liu, Qingshan, Bibi, Adel, King, Irwin, Lyu, Siwei
Format:	Preprint
Published:	2026
Subjects:	Cryptography and Security Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Red-Teaming Agent Execution Contexts: Open-World Security Evaluation on OpenClaw
by: Yao, Hongwei, et al.
Published: (2026)

Can ChatGPT Detect DeepFakes? A Study of Using Multimodal Large Language Models for Media Forensics
by: Jia, Shan, et al.
Published: (2024)

Analysing Safety Risks in LLMs Fine-Tuned with Pseudo-Malicious Cyber Security Data
by: ElZemity, Adel, et al.
Published: (2025)

FinVault: Benchmarking Financial Agent Safety in Execution-Grounded Environments
by: Yang, Zhi, et al.
Published: (2026)

CapSeal: Capability-Sealed Secret Mediation for Secure Agent Execution
by: Jin, Shutong, et al.
Published: (2026)

CyberLLMInstruct: A Pseudo-malicious Dataset Revealing Safety-performance Trade-offs in Cyber Security LLM Fine-tuning
by: ElZemity, Adel, et al.
Published: (2025)

Agent Security is a Systems Problem
by: Christodorescu, Mihai, et al.
Published: (2026)

Your Agent, Their Asset: A Real-World Safety Analysis of OpenClaw
by: Wang, Zijun, et al.
Published: (2026)

Measuring Safety Alignment Effects in Autonomous Security Agents
by: David, Isaac, et al.
Published: (2026)

Towards Unifying Quantitative Security Benchmarking for Multi Agent Systems
by: Sharma, Gauri, et al.
Published: (2025)

ToolTweak: An Attack on Tool Selection in LLM-based Agents
by: Sneh, Jonathan, et al.
Published: (2025)

Autonomous Intelligent Agents for Natural-Language-Driven Web Execution with Integrated Security Assurance
by: Pasupuleti, Vinil, et al.
Published: (2026)

Attacks in Adversarial Machine Learning: A Systematic Survey from the Life-cycle Perspective
by: Wu, Baoyuan, et al.
Published: (2023)

Security Risks in Tool-Enabled AI Agents: A Systematic Analysis of Privileged Execution Environments
by: Goel, Hardik
Published: (2026)

A Security Analysis of the OpenClaw AI Agent Framework
by: Suwansathit, Surada, et al.
Published: (2026)

Taming OpenClaw: Security Analysis and Mitigation of Autonomous LLM Agent Threats
by: Deng, Xinhao, et al.
Published: (2026)

Class-Conditional Neural Polarizer: A Lightweight and Effective Backdoor Defense by Purifying Poisoned Features
by: Zhu, Mingli, et al.
Published: (2025)

ClawKeeper: Comprehensive Safety Protection for OpenClaw Agents Through Skills, Plugins, and Watchers
by: Liu, Songyang, et al.
Published: (2026)

Provably Secure Agent Guardrail
by: Wu, Benlong, et al.
Published: (2026)

Adaptive Plan-Execute Framework for Smart Contract Security Auditing
by: Wei, Zhiyuan, et al.
Published: (2025)

Agent-Sentry: Bounding LLM Agents via Execution Provenance
by: Sequeira, Rohan, et al.
Published: (2026)

Do Coding Agents Understand Least-Privilege Authorization?
by: Yan, Zheng, et al.
Published: (2026)

OpenPort Protocol: A Security Governance Specification for AI Agent Tool Access
by: Zhu, Genliang, et al.
Published: (2026)

Towards Secure Agent Skills: Architecture, Threat Taxonomy, and Security Analysis
by: Li, Zhiyuan, et al.
Published: (2026)

SoK: Trust-Authorization Mismatch in LLM Agent Interactions
by: Shi, Guanquan, et al.
Published: (2025)

APEX: Agent Payment Execution with Policy for Autonomous Agent API Access
by: Uddin, Mohd Safwan, et al.
Published: (2026)

OSS-CRS: Liberating AIxCC Cyber Reasoning Systems for Real-World Open-Source Security
by: Chin, Andrew, et al.
Published: (2026)

Open Challenges in Multi-Agent Security: Towards Secure Systems of Interacting AI Agents
by: de Witt, Christian Schroeder, et al.
Published: (2025)

AgentWard: A Lifecycle Security Architecture for Autonomous AI Agents
by: Zhang, Yixiang, et al.
Published: (2026)

SecRepoBench: Benchmarking Code Agents for Secure Code Completion in Real-World Repositories
by: Shen, Chihao, et al.
Published: (2025)

RAS-Eval: A Comprehensive Benchmark for Security Evaluation of LLM Agents in Real-World Environments
by: Fu, Yuchuan, et al.
Published: (2025)

Architecting Resilient LLM Agents: A Guide to Secure Plan-then-Execute Implementations
by: Del Rosario, Ron F., et al.
Published: (2025)

ClawTrap: A MITM-Based Red-Teaming Framework for Real-World OpenClaw Security Evaluation
by: Zhao, Haochen, et al.
Published: (2026)

Fragile Model Watermark for integrity protection: leveraging boundary volatility and sensitive sample-pairing
by: Gao, ZhenZhe, et al.
Published: (2024)

AIRGuard: Guarding Agent Actions with Runtime Authority Control
by: Qin, Suliu, et al.
Published: (2026)

A New Era in LLM Security: Exploring Security Concerns in Real-World LLM-based Systems
by: Wu, Fangzhou, et al.
Published: (2024)

Quantifying Security Vulnerabilities: A Metric-Driven Security Analysis of Gaps in Current AI Standards
by: Madhavan, Keerthana, et al.
Published: (2025)

Security of AI Agents
by: He, Yifeng, et al.
Published: (2024)

A Safety and Security Framework for Real-World Agentic Systems
by: Ghosh, Shaona, et al.
Published: (2025)

Progent: Securing AI Agents with Privilege Control
by: Shi, Tianneng, et al.
Published: (2025)