Fetching the paper…

Reading the bibliography…

2025

A Framework for Evaluating Emerging Cyberattack Capabilities of AI

Rodriguez, Mikel, Popa, Raluca Ada, Flynn, Four et al.

Original
PDF
Post on X

Understand

As frontier AI models become more capable, evaluating their potential to enable cyberattacks is crucial for ensuring the safe development of Artificial General Intelligence (AGI).

Current cyber evaluation efforts are often ad-hoc, lacking systematic analysis of attack phases and guidance on targeted defenses.
This work introduces a novel evaluation framework that addresses these limitations by: (1) examining the end-to-end attack chain, (2) identifying gaps in AI threat evaluation, and (3) helping defenders prioritize targeted mitigations and conduct AI-enabled adversary emulation for red teaming.
Our approach adapts existing cyberattack chain frameworks for AI systems.

A Framework for Evaluating Emerging Cyberattack Capabilities of AI · Around

Built on

Zero days, thousands of nights
L. Ablon and A. Bogart · 2017
Earlier work this paper cites.
Vuldeepecker: A deep learning-based system for vulnerability detection
Original
Z. Li, D. Zou, S. Xu, X. Ou, H. Jin, S. Wang, Z. Deng, and Y. Zhong · 2018
Earlier work this paper cites.
Mitre att&ck: Design and philosophy
B. E. Strom, A. Applebaum, D. P. Miller, K. C. Nickels, A. G. Pennington, and C. B. Thomas · 2018
Earlier work this paper cites.
Threat alert prioritization using isolation forest and stacked auto encoder with day-forward-chaining analysis
M. E. Aminanto, T. Ban, R. Isawa, T. Takahashi, and D. Inoue · 2020
Earlier work this paper cites.
Breaking the cyber kill chain by modelling resource costs
K. Haga, P. H. Meland, and G. Sindre · 2020
Earlier work this paper cites.
Sysevr: A framework for using deep learning to detect software vulnerabilities
Z. Li, D. Zou, S. Xu, H. Jin, Y. Zhu, and Z. Chen · 2021
Earlier work this paper cites.
Vulnerability detection and monitoring using llm
V. Akuthota, R. Kasula, S. T. Sumona, M. Mohiuddin, M. T. Reza, and M. M. Rahman · 2023
Earlier work this paper cites.
Breaking alert fatigue: Ai-assisted siem framework for effective incident response
T. Ban, T. Takahashi, S. Ndichu, and D. Inoue · 2023
Earlier work this paper cites.
Purple llama cyberseceval: A secure coding benchmark for language models
Original
M. Bhatt, S. Chennabasappa, C. Nikolaidis, S. Wan, I. Evtimov, D. Gabi, D. Song, F. Ahmad, C. Aschermann, L. Fontana, et al · 2023
Earlier work this paper cites.
Pentestgpt: An llm-empowered automatic penetration testing tool
Original
G. Deng, Y. Liu, V. Mayoral-Vilches, P. Liu, Y. Li, Y. Xu, T. Zhang, Y. Liu, M. Pinzger, and S. Rass · 2023
Earlier work this paper cites.
Penheal: A two-stage llm framework for automated pentesting and optimal remediation
J. Huang and Q. Zhu · 2023
Earlier work this paper cites.
Understanding the effectiveness of large language models in detecting security vulnerabilities
Original
A. Khare, S. Dutta, Z. Li, A. Solko-Breslin, R. Alur, and M. Naik · 2023
Earlier work this paper cites.
Secqa: A concise question-answering dataset for evaluating large language models in computer security
Original
Z. Liu · 2023
Earlier work this paper cites.
Model evaluation for extreme risks
Original
T. Shevlane, S. Farquhar, B. Garfinkel, M. Phuong, J. Whittlestone, J. Leung, D. Kokotajlo, N. Marchal, M. Anderljung, N. Kolt, et al · 2023
Earlier work this paper cites.
Using large language models for cybersecurity capture-the-flag challenges and certification questions. arxiv 2023
Original
W. Tann, Y. Liu, J. Sim, C. Seah, and E. Chang · 2023
Earlier work this paper cites.

Similar

Enigma: Enhanced interactive generative model agent for ctf challenges
Original
T. Abramovich, M. Udeshi, M. Shao, K. Lieret, H. Xi, K. Milner, S. Jancheska, J. Yang, C. E. Jimenez, F. Khorrami, et al · 2024
Cited alongside, same era.
Exploring llms for malware detection: Review, framework design, and countermeasure approaches
Original
J. Al-Karaki, M. A.-Z. Khan, and M. Omar · 2024
Cited alongside, same era.
Ctibench: A benchmark for evaluating llms in cyber threat intelligence
Original
M. T. Alam, D. Bhusal, L. Nguyen, and N. Rastogi · 2024
Cited alongside, same era.
Cyberseceval 2: A wide-ranging cybersecurity evaluation suite for large language models
Original
M. Bhatt, S. Chennabasappa, Y. Li, C. Nikolaidis, D. Song, S. Wan, F. Ahmad, C. Aschermann, Y. Chen, D. Kapil, et al · 2024
Cited alongside, same era.

Then

Nyu ctf bench: A scalable open-source benchmark dataset for evaluating llms in offensive security
M. Shao, S. Jancheska, M. Udeshi, B. Dolan-Gavitt, K. Milner, B. Chen, M. Yin, S. Garg, P. Krishnamurthy, F. Khorrami, et al · 2024
Later among the works it cites.
Cybermetric: A benchmark dataset for evaluating large language models knowledge in cybersecurity
Original
N. Tihanyi, M. A. Ferrag, R. Jain, and M. Debbah · 2024
Later among the works it cites.
Cyberseceval 3: Advancing the evaluation of cybersecurity risks and capabilities in large language models
Original
S. Wan, C. Nikolaidis, D. Song, D. Molnar, J. Crnkovich, J. Grace, M. Bhatt, S. Chennabasappa, S. Whitman, S. Ding, et al · 2024
Later among the works it cites.
Autoattacker: A large language model guided system to implement automatic cyber-attacks
Original
J. Xu, J. W. Stokes, G. McDonald, X. Bai, D. Marshall, S. Wang, A. Swaminathan, and Z. Li · 2024
Later among the works it cites.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…

garak: A framework for security probing large language models

L. Derczynski, E. Galinkin, J. Martin, S. Majumdar, and N. Inie · 2024

Cited alongside, same era.

Vul-rag: Enhancing llm-based vulnerability detection via knowledge-level rag

X. Du, G. Zheng, K. Wang, J. Feng, W. Deng, M. Liu, B. Chen, X. Peng, T. Ma, and Y. Lou · 2024

Cited alongside, same era.

Safety case template for frontier ai: A cyber inability argument

A. Goemans, M. D. Buhl, J. Schuett, T. Korbak, J. Wang, B. Hilton, and G. Irving · 2024

Cited alongside, same era.

Project naptime: Evaluating offensive security capabilities of large language models, 2025c

Google · 2024

Cited alongside, same era.

Employing llms for incident response planning and review

S. Hays and J. White · 2024

Cited alongside, same era.

Openai o1 system card

A. Jaech, A. Kalai, A. Lerer, A. Richardson, A. El-Kishky, A. Low, A. Helyar, A. Madry, A. Beutel, A. Carney, et al · 2024

Cited alongside, same era.

Perception shapes reality: How views on financial market correlation affect capital availability for cyber insurance

T. Johansmeyer · 2024

Cited alongside, same era.

Grace: Empowering llm-based software vulnerability detection with graph structure and in-context learning

G. Lu, X. Ju, X. Chen, W. Pei, and Z. Cai · 2024

Cited alongside, same era.

Swe-agent: Agent-computer interfaces enable automated software engineering

J. Yang, C. Jimenez, A. Wettig, K. Lieret, S. Yao, K. Narasimhan, and O. Press · 2024

Later among the works it cites.

Cybench: A framework for evaluating cybersecurity capabilities and risks of language models

A. K. Zhang, N. Perry, R. Dulepet, J. Ji, C. Menders, J. W. Lin, E. Jones, G. Hussein, S. Liu, D. Jasper, et al · 2024

Later among the works it cites.

Claude 3.7 sonnet system card, 2025

Anthropic · 2025

Closest in time.

Cyber events database, 2025

Center for Strategic and International Studies · 2025

Closest in time.

Aixcc: Ai cyber challenge, 2025

DARPA · 2025

Closest in time.

Frontier safety framework 2.0, 2025

Google DeepMind · 2025

Closest in time.

Uk ai safety institute rebrands, 2025

InfoSecurity Magazine · 2025

Closest in time.

Occult: Evaluating large language models for offensive cyber operation capabilities

M. Kouremetis, M. Dotter, A. Byrne, D. Martin, E. Michalak, G. Russo, M. Threet, and G. Zarrella · 2025

Closest in time.

Cyber kill chain, 2025

Lockheed Martin · 2025

Closest in time.

Mandiant advantage: Threat intelligence, 2025

Mandiant · 2025

Closest in time.

On the feasibility of using llms to execute multistage network attacks

Original

B. Singer, K. Lucas, L. Adiga, M. Jain, L. Bauer, and V. Sekar · 2025

Closest in time.