智能体AI的黑盒红队测试：基于分类学的自动化风险发现框架

中文摘要 AI 辅助

智能体系统正迅速投入生产环境，它们读取不可信输入、调用具有真实权限的工具并自主行动，将安全边界扩展到了仅限聊天的模型之外。然而，标准评估仍停留在单轮交互层面，无法捕捉多步骤的智能体漏洞。我们提出了一个系统的黑盒框架，用于风险感知的智能体评估，该框架仅需基本的系统描述。我们的方法引入了：（1）一个七域分类学，将可观察行为映射到风险类别；（2）完全自动化的SAGE-RT红队测试，为每个域生成120个对抗性场景；（3）使用LLM评判员进行人工验证的评估。在两种智能体架构（CrewAI和AutoGen）上使用四个基础模型进行的实证验证揭示了令人担忧的模式：平均治理风险为56.25%，多智能体配置中的隐私风险为65%，智能体行为漏洞高达85%。我们的黑盒方法无需特权访问即可有效识别关键架构漏洞，为更安全的智能体部署提供了一条可扩展的路径。

英文摘要

Agentic systems are rapidly moving to production, where they read untrusted inputs, call tools with real permissions, and act autonomously, expanding the security surface beyond chat-only models. Yet standard evaluations remain single-turn and fail to capture multi-step agent vulnerabilities. We present a systematic black-box framework for risk-aware agent evaluation requiring only basic system descriptions. Our approach introduces: (1) a seven-domain taxonomy mapping observable behaviors to risk categories, (2) fully automated SAGE-RT red teaming producing 120 adversarial scenarios per domain, and (3) human-validated evaluation using LLM judges. Empirical validation across two agent architectures (CrewAI and AutoGen) with four base models reveals alarming patterns: 56.25\% average governance risk, 65\% privacy risk in multi-agent configurations, and agent behavior vulnerabilities reaching 85\%. Our black-box approach effectively identifies critical architectural vulnerabilities without privileged access, providing a scalable path toward safer agent deployments.

发表机构

• Enkrypt AI

机构由 AI 辅助整理，请以论文原文为准。