A Generative Artificial Intelligence Methodology for Automated Zero-Shot Data Tagging to Support Tactical Zero Trust Architecture Implementation
DOI:
https://doi.org/10.37266/ISER.2025v12i2.pp83-88Keywords:
Generative AI, Large Language Model, In-Context Learning, Data Tagging, Zero Trust Architecture, CybersecurityAbstract
A significant challenge in the implementation of the military’s cybersecurity framework for Zero Trust Architecture (ZTA) is that the current approach for data tagging is done manually, which is a time-consuming and error-prone process that undermines the efficiency and effectiveness of cybersecurity measures. This paper introduces an innovative methodology that leverages generative artificial intelligence (AI) for automated data tagging to support tactical ZTA implementation within mili- tary mission command systems (MCS). Specifically, we develop the Generative AI ZEro-trust Labeling (GAZEL) tool, which uses a fine-tuned Large Language Model combined with in-context learning for automated zero-shot tagging of MCS message data according to predefined access control categories, streamlining the path towards an agile and fortified cybersecurity posture.
References
Army Cyber Center of Excellence. (2023). Zero trust tactical implementation guide (Technical Report). Fort Gordon, Georgia: Cyber Center of Excellence, Fort Gordon. (Controlled by: Army CCoE. CUI Category: CTI, OPSEC. Dissemination Control: FEDCON.)
Barrett, C., Boyd, B., Bursztein, E., Carlini, N., Chen, B., Choi, J., … Yang, D. (2023). Identifying and mitigating the security risks of generative ai. Foundations and Trends® in Privacy and Security, 6(1), 1–52. doi: 10.1561/3300000041
Bommasani, R., Hudson, D. A., Adeli, E., Altman, R., Arora, S., von Arx, S., … Liang, P. (2021). On the opportunities and risks of foundation models. ArXiv, abs/2108.07258. Retrieved from https://api.semanticscholar.org/CorpusID: 237091588
Cybersecurity and Infrastructure Security Agency. (2020). Ed 21-01: Mitigate solarwinds orion code compromise. Retrieved from https://www.cisa.gov/news-events/directives/ed-21-01-mitigate-solarwinds-orion-code-compromise.
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019). Bert: Pre-training of deep bidirectional transformers for language understanding.
Ding, B., Qin, C., Liu, L., Chia, Y. K., Joty, S., Li, B., & Bing, L. (2023). Is gpt-3 a good data annotator?
Executive order 14028: Improving the nation’s cybersecurity. (2021). Retrieved from https://www.gsa.gov/technology/ it-contract-vehicles-and-purchasing-programs/technology-products-services/it-security/executive-order-14028
Goodfellow, I. J., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., … Bengio, Y. (2014). Generative adversarial nets. In Proceedings of the 27th international conference on neural information processing systems - volume 2 (p. 2672– 2680). Cambridge, MA, USA: MIT Press.
Hu, J. E., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., & Chen, W. (2021). Lora: Low-rank adaptation of large language models. ArXiv, abs/2106.09685. Retrieved from https://api.semanticscholar.org/CorpusID:235458009
ISOO. (2018). Marking classified national security information. Retrieved from https://www.archives.gov/files/ isoo/training/marking-booklet-revision.pdf
Kaur, R., Gabrijelčič, D., & Klobučar, T. (2023). Artificial intelligence for cybersecurity: Literature review and future research directions. Information Fusion, 97, 101804. doi: https://doi.org/10.1016/j.inffus.2023.101804
Keall, C. (2019). Experts warning as microsoft restores office 365 after worldwide outage. Retrieved from
https://www.nzherald.co.nz/business/experts-warning-as-microsoft-restores-office-365-after
-worldwide-outage/JZVIHQJZSLT3WPXTEAXBVROSGQ/?c_id=3&objectid=12286870
Kucharavy, A., Schillaci, Z. M., Mar’echal, L., Wursch, M., Dolamic, L., Sabonnadiere, R., … Lenders, V. (2023). Funda- mentals of generative large language models and perspectives in cyber-defense. ArXiv, abs/2303.12132. Retrieved from https://api.semanticscholar.org/CorpusID:257663521
National Institute of Standards and Technology. (2018). Framework for improving critical infrastructure cybersecurity. Retrieved from https://nvlpubs.nist.gov/nistpubs/cswp/nist.cswp.04162018.pdf
Pangakis, N., Wolken, S., & Fasching, N. (2023). Automated annotation with generative ai requires validation. ArXiv, abs/2306.00176. Retrieved from https://api.semanticscholar.org/CorpusID:259000016
Pourpanah, F., Abdar, M., Luo, Y., Zhou, X., Wang, R., Lim, C. P., … Wu, Q. M. J. (2023). A review of generalized zero-shot learning methods. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(4), 4051-4070.
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., … Polosukhin, I. (2017). Atten- tion is all you need. In I. Guyon et al. (Eds.), Advances in neural information processing systems (Vol. 30). Curran Associates, Inc. Retrieved from https://proceedings.neurips.cc/paper_files/paper/2017/file/ 3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf
Wolfram, S. (2023, Feb). What is chatgpt doing ... and why does it work? Retrieved from https://writings.stephenwolfram.com/2023/02/what-is-chatgpt-doing-and-why-does-it-work/ (Accessed: 2023-03-08)
Woodiss-Field, A., Johnstone, M. N., & Haskell-Dowland, P. (2024). Examination of traditional botnet detection on iot-based bots. Sensors, 24(3). Retrieved from https://www.mdpi.com/1424-8220/24/3/1027 doi: 10.3390/s24031027
Young, S. D. (2021). Improving detection of cybersecurity vulnerabilities and incidents on federal government systems through endpoint detection and response. Retrieved from https://www.whitehouse.gov/wp-content/uploads/2021/10/ M-22-01.pdf
Young, S. D. (2022). Moving the u.s. government toward zero trust cybersecurity principles. Retrieved from https:// www.whitehouse.gov/wp-content/uploads/2022/01/M-22-09.pdf
Zhang, P., Zeng, G., Wang, T., & Lu, W. (2024). TinyLlama: An open-source small language model.
Published
How to Cite
Issue
Section
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
The copyediting stage is intended to improve the flow, clarity, grammar, wording, and formatting of the article. It represents the last chance for the author to make any substantial changes to the text because the next stage is restricted to typos and formatting corrections. The file to be copyedited is in Word or .rtf format and therefore can easily be edited as a word processing document. The set of instructions displayed here proposes two approaches to copyediting. One is based on Microsoft Word's Track Changes feature and requires that the copy editor, editor, and author have access to this program. A second system, which is software independent, has been borrowed, with permission, from the Harvard Educational Review. The journal editor is in a position to modify these instructions, so suggestions can be made to improve the process for this journal.
