Evaluating Privilege Usage of Agents with Real-World Tools

TL;DR

This paper introduces GrantBox, a sandbox for evaluating LLM agents' privilege control with real-world tools, revealing vulnerabilities to sophisticated prompt injection attacks.

Contribution

The paper presents GrantBox, a novel evaluation environment that integrates real-world tools for assessing privilege usage and security of LLM agents.

Findings

LLMs can block some direct privilege attacks.

Agents remain vulnerable to sophisticated prompt injection attacks.

Average attack success rate was 84.80% in tested scenarios.

Abstract

Equipping LLM agents with real-world tools can substantially improve productivity. However, granting agents autonomy over tool use also transfers the associated privileges to both the agent and the underlying LLM. Improper privilege usage may lead to serious consequences, including information leakage and infrastructure damage. While several benchmarks have been built to study agents' security, they often rely on pre-coded tools and restricted interaction patterns. Such crafted environments differ substantially from the real-world, making it hard to assess agents' security capabilities in critical privilege control and usage. Therefore, we propose GrantBox, a security evaluation sandbox for analyzing agent privilege usage. GrantBox automatically integrates real-world tools and allows LLM agents to invoke genuine privileges, enabling the evaluation of privilege usage under prompt injection attacks. Our results indicate that while LLMs exhibit basic security awareness and can block some direct attacks, they remain vulnerable to more sophisticated attacks, resulting in an average attack success rate of 84.80% in carefully crafted scenarios.