Copertina

Evaluating Privilege Usage of Agents with Real-World Tools

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Zhang, Quan, Fu, Lianhang, Lian, Lvsi, Go, Gwihwan, Wang, Yujue, Zhou, Chijin, Jiang, Yu, Pu, Geguang
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Cryptography and Security Artificial Intelligence
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Human-Imperceptible Retrieval Poisoning Attacks in LLM-Powered Applications
di: Zhang, Quan, et al.
Pubblicazione: (2024)

MiniScope: A Least Privilege Framework for Authorizing Tool Calling Agents
di: Zhu, Jinhao, et al.
Pubblicazione: (2025)

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

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

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

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

LSPRAG: LSP-Guided RAG for Language-Agnostic Real-Time Unit Test Generation
di: Go, Gwihwan, et al.
Pubblicazione: (2025)

From Controlled to the Wild: Evaluation of Pentesting Agents for the Real-World
di: Conde, Pedro, et al.
Pubblicazione: (2026)

BashArena: A Control Setting for Highly Privileged AI Agents
di: Kaufman, Adam, et al.
Pubblicazione: (2025)

Enhancing Linux Privilege Escalation Attack Capabilities of Local LLM Agents
di: Probst, Benjamin, et al.
Pubblicazione: (2026)

Post-Training Local LLM Agents for Linux Privilege Escalation with Verifiable Rewards
di: Normann, Philipp, et al.
Pubblicazione: (2026)

We Urgently Need Privilege Management in MCP: A Measurement of API Usage in MCP Ecosystems
di: Li, Zhihao, et al.
Pubblicazione: (2025)

Agent Privilege Separation in OpenClaw: A Structural Defense Against Prompt Injection
di: Cheng, Darren, et al.
Pubblicazione: (2026)

NCCR: to Evaluate the Robustness of Neural Networks and Adversarial Examples
di: Pu, Shi, et al.
Pubblicazione: (2025)

AgentTrust: Runtime Safety Evaluation and Interception for AI Agent Tool Use
di: Yang, Chenglin
Pubblicazione: (2026)

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

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

LLMs as Hackers: Autonomous Linux Privilege Escalation Attacks
di: Happe, Andreas, et al.
Pubblicazione: (2023)

Prompt Flow Integrity to Prevent Privilege Escalation in LLM Agents
di: Kim, Juhee, et al.
Pubblicazione: (2025)

AgentGuard: Repurposing Agentic Orchestrator for Safety Evaluation of Tool Orchestration
di: Chen, Jizhou, et al.
Pubblicazione: (2025)

ChainCaps: Composition-Safe Tool-Using Agents via Monotonic Capability Attenuation
di: Jiang, Xiaochong, et al.
Pubblicazione: (2026)

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

CyberGym: Evaluating AI Agents' Real-World Cybersecurity Capabilities at Scale
di: Wang, Zhun, et al.
Pubblicazione: (2025)

AdInject: Real-World Black-Box Attacks on Web Agents via Advertising Delivery
di: Wang, Haowei, et al.
Pubblicazione: (2025)

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

Automating Security Audit Using Large Language Model based Agent: An Exploration Experiment
di: Chin, Jia Hui, et al.
Pubblicazione: (2025)

VIGIL: Defending LLM Agents Against Tool Stream Injection via Verify-Before-Commit
di: Lin, Junda, et al.
Pubblicazione: (2026)

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

LogicEval: A Systematic Framework for Evaluating Automated Repair Techniques for Logical Vulnerabilities in Real-World Software
di: Rashid, Syed Md Mukit, et al.
Pubblicazione: (2026)

Beyond Max Tokens: Stealthy Resource Amplification via Tool Calling Chains in LLM Agents
di: Zhou, Kaiyu, et al.
Pubblicazione: (2026)

Certified Causal Attribution for Real-Time Attack Forensics in 6G Network Slicing
di: Quan, Minh K., et al.
Pubblicazione: (2026)

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

Red-Teaming Coding Agents from a Tool-Invocation Perspective: An Empirical Security Assessment
di: Xie, Yuchong, et al.
Pubblicazione: (2025)

Domain-Adapted Granger Causality for Real-Time Cross-Slice Attack Attribution in 6G Networks
di: Quan, Minh K., et al.
Pubblicazione: (2025)

BackdoorMBTI: A Backdoor Learning Multimodal Benchmark Tool Kit for Backdoor Defense Evaluation
di: Yu, Haiyang, et al.
Pubblicazione: (2024)

Causality Laundering: Denial-Feedback Leakage in Tool-Calling LLM Agents
di: Chinaei, Mohammad Hossein
Pubblicazione: (2026)

Free-MAD: Consensus-Free Multi-Agent Debate
di: Cui, Yu, et al.
Pubblicazione: (2025)

Comparing AI Agents to Cybersecurity Professionals in Real-World Penetration Testing
di: Lin, Justin W., et al.
Pubblicazione: (2025)

A Comparative Evaluation of AI Agent Security Guardrails
di: Li, Qi, et al.
Pubblicazione: (2026)

Before the Tool Call: Deterministic Pre-Action Authorization for Autonomous AI Agents
di: Uchibeke, Uchi
Pubblicazione: (2026)