icon-carat-right menu search cmu-wordmark

What Can Generative AI Red-Teaming Learn from Cyber Red-Teaming?

Technical Report
This paper investigates the applicability of established cyber red-teaming methodologies to the evaluation of generative AI systems, addressing the growing need for robust security assessments in AI-driven applications.
Publisher

Software Engineering Institute

CMU/SEI Report Number
CMU/SEI-2025-TR-006
DOI (Digital Object Identifier)
10.1184/R1/29410136

Abstract

Red teaming, a security practice rooted in adversarial emulation, has been widely applied across various domains, including cybersecurity and artificial intelligence (AI). This paper investigates the applicability of established cyber red-teaming methodologies to the evaluation of generative AI systems, addressing the growing need for robust security assessments in AI-driven applications. Through a pair of systematic literature reviews, we synthesize existing generative AI red-teaming approaches and analyze their alignment with established practices in cyber red-teaming.

Cite This Technical Report

Sinha, A., Lucassen, J., Grimes, K., Feffer, M., Soto, M., Heidari, H., & VanHoudnos, N. (2025, July 16). What Can Generative AI Red-Teaming Learn from Cyber Red-Teaming?. (Technical Report CMU/SEI-2025-TR-006). Retrieved August 11, 2025, from https://doi.org/10.1184/R1/29410136.

@techreport{sinha_2025,
author={Sinha, Anusha and Lucassen, James and Grimes, Keltin and Feffer, Michael and Soto, Mary and Heidari, Hoda and VanHoudnos, Nathan},
title={What Can Generative AI Red-Teaming Learn from Cyber Red-Teaming?},
month={{Jul},
year={{2025},
number={{CMU/SEI-2025-TR-006},
howpublished={Carnegie Mellon University, Software Engineering Institute's Digital Library},
url={https://doi.org/10.1184/R1/29410136},
note={Accessed: 2025-Aug-11}
}

Sinha, Anusha, James Lucassen, Keltin Grimes, Michael Feffer, Mary Soto, Hoda Heidari, and Nathan VanHoudnos. "What Can Generative AI Red-Teaming Learn from Cyber Red-Teaming?." (CMU/SEI-2025-TR-006). Carnegie Mellon University, Software Engineering Institute's Digital Library. Software Engineering Institute, July 16, 2025. https://doi.org/10.1184/R1/29410136.

A. Sinha, J. Lucassen, K. Grimes, M. Feffer, M. Soto, H. Heidari, and N. VanHoudnos, "What Can Generative AI Red-Teaming Learn from Cyber Red-Teaming?," Carnegie Mellon University, Software Engineering Institute's Digital Library. Software Engineering Institute, Technical Report CMU/SEI-2025-TR-006, 16-Jul-2025 [Online]. Available: https://doi.org/10.1184/R1/29410136. [Accessed: 11-Aug-2025].

Sinha, Anusha, James Lucassen, Keltin Grimes, Michael Feffer, Mary Soto, Hoda Heidari, and Nathan VanHoudnos. "What Can Generative AI Red-Teaming Learn from Cyber Red-Teaming?." (Technical Report CMU/SEI-2025-TR-006). Carnegie Mellon University, Software Engineering Institute's Digital Library, Software Engineering Institute, 16 Jul. 2025. https://doi.org/10.1184/R1/29410136. Accessed 11 Aug. 2025.

Sinha, Anusha; Lucassen, James; Grimes, Keltin; Feffer, Michael; Soto, Mary; Heidari, Hoda; & VanHoudnos, Nathan. What Can Generative AI Red-Teaming Learn from Cyber Red-Teaming?. CMU/SEI-2025-TR-006. Software Engineering Institute. 2025. https://doi.org/10.1184/R1/29410136