MATRA: Modeling the Attack Surface of Agentic AI Systems -- OpenClaw Case Study
Tim Van hamme, Thomas Vissers, Javier Carnerero-Cano, Mario Fritz, Emil C. Lupu + 2 more
TLDR
MATRA is a threat modeling framework for agentic AI systems, assessing risks and quantifying how architectural controls reduce attack blast radius.
Key contributions
- Introduces MATRA, a pragmatic threat modeling framework for agentic AI systems.
- Adapts established risk assessment methods, using asset-based impact and attack trees.
- Demonstrates MATRA on an OpenClaw personal AI agent, identifying deployment-specific risks.
- Quantifies risk reduction from architectural controls like sandboxing and least-privilege access.
Why it matters
This paper addresses a critical gap in securing autonomous AI agents by providing a systematic threat modeling framework. It helps practitioners assess and mitigate risks, demonstrating how architectural controls effectively limit the impact of attacks. This is crucial for safe and responsible AI deployment.
Original Abstract
LLMs are increasingly deployed as autonomous agents with access to tools, databases, and external services, yet practitioners (across different sectors) lack systematic methods to assess how known threat classes translate into concrete risks within a specific agentic deployment. We present MATRA, a pragmatic threat modeling framework for agentic AI systems that adapts established risk assessment methodology to systematically assess how known LLM threats translate into deployment-specific risks. MATRA begins with an asset-based impact assessment and utilizes attack trees to determine the likelihood of these impacts occurring within the system architecture. We demonstrate MATRA on a personal AI agent deployment using OpenClaw, quantifying how architectural controls such as network sandboxing and least-privilege access reduce risk by limiting the blast radius of successful injections.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.