DoomArena: A framework for Testing AI Agents Against Evolving Security Threats
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866912632894652416 |
|---|---|
| author | Boisvert, Leo Bansal, Mihir Evuru, Chandra Kiran Reddy Huang, Gabriel Puri, Abhay Bose, Avinandan Fazel, Maryam Cappart, Quentin Stanley, Jason Lacoste, Alexandre Drouin, Alexandre Dvijotham, Krishnamurthy |
| author_facet | Boisvert, Leo Bansal, Mihir Evuru, Chandra Kiran Reddy Huang, Gabriel Puri, Abhay Bose, Avinandan Fazel, Maryam Cappart, Quentin Stanley, Jason Lacoste, Alexandre Drouin, Alexandre Dvijotham, Krishnamurthy |
| contents | We present DoomArena, a security evaluation framework for AI agents. DoomArena is designed on three principles: 1) It is a plug-in framework and integrates easily into realistic agentic frameworks like BrowserGym (for web agents) and $τ$-bench (for tool calling agents); 2) It is configurable and allows for detailed threat modeling, allowing configuration of specific components of the agentic framework being attackable, and specifying targets for the attacker; and 3) It is modular and decouples the development of attacks from details of the environment in which the agent is deployed, allowing for the same attacks to be applied across multiple environments. We illustrate several advantages of our framework, including the ability to adapt to new threat models and environments easily, the ability to easily combine several previously published attacks to enable comprehensive and fine-grained security testing, and the ability to analyze trade-offs between various vulnerabilities and performance. We apply DoomArena to state-of-the-art (SOTA) web and tool-calling agents and find a number of surprising results: 1) SOTA agents have varying levels of vulnerability to different threat models (malicious user vs malicious environment), and there is no Pareto dominant agent across all threat models; 2) When multiple attacks are applied to an agent, they often combine constructively; 3) Guardrail model-based defenses seem to fail, while defenses based on powerful SOTA LLMs work better. DoomArena is available at https://github.com/ServiceNow/DoomArena. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_14064 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DoomArena: A framework for Testing AI Agents Against Evolving Security Threats Boisvert, Leo Bansal, Mihir Evuru, Chandra Kiran Reddy Huang, Gabriel Puri, Abhay Bose, Avinandan Fazel, Maryam Cappart, Quentin Stanley, Jason Lacoste, Alexandre Drouin, Alexandre Dvijotham, Krishnamurthy Cryptography and Security We present DoomArena, a security evaluation framework for AI agents. DoomArena is designed on three principles: 1) It is a plug-in framework and integrates easily into realistic agentic frameworks like BrowserGym (for web agents) and $τ$-bench (for tool calling agents); 2) It is configurable and allows for detailed threat modeling, allowing configuration of specific components of the agentic framework being attackable, and specifying targets for the attacker; and 3) It is modular and decouples the development of attacks from details of the environment in which the agent is deployed, allowing for the same attacks to be applied across multiple environments. We illustrate several advantages of our framework, including the ability to adapt to new threat models and environments easily, the ability to easily combine several previously published attacks to enable comprehensive and fine-grained security testing, and the ability to analyze trade-offs between various vulnerabilities and performance. We apply DoomArena to state-of-the-art (SOTA) web and tool-calling agents and find a number of surprising results: 1) SOTA agents have varying levels of vulnerability to different threat models (malicious user vs malicious environment), and there is no Pareto dominant agent across all threat models; 2) When multiple attacks are applied to an agent, they often combine constructively; 3) Guardrail model-based defenses seem to fail, while defenses based on powerful SOTA LLMs work better. DoomArena is available at https://github.com/ServiceNow/DoomArena. |
| title | DoomArena: A framework for Testing AI Agents Against Evolving Security Threats |
| topic | Cryptography and Security |
| url | https://arxiv.org/abs/2504.14064 |