Machine Learning-Based Run-Time DevSecOps: ChatGPT Against Traditional Approach
Abstract
DevSecOps paradigm strives to provide the extension of the already approved DevOps practice by tackling the issues related to security in both design-and run-time. This paper focuses on aspects of run-time security and proposes a server log analysis-based machine learning-empowered method. The goal is to detect suspicious activity based on traffic-related information. For this purpose, we propose a novel dialogue-alike approach relying on ChatGPT in Python, where context represents labelled data, while the questions themselves contain the log records which are about to be evaluated if they represent suspicious activity or not. On the otter side, a traditional machine learning approach treating the problem as classification using Weka API in Java is also presented, for comparison. The evaluation was performed on publicly available server log dataset from Kaggle. According to the achieved results, ChatGPT-based approach shows potential to be used for run-time DevSecOps scenarios, but still has many limitations and drawbacks (limited context, processing time, costs, lower accuracy) while it is still not as effective as traditional, already approved classification techniques.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Findings indicate improved real-time security integration, reduced manual intervention, and increased operational resilience in DevSecOps pipelines.
“Findings indicate improved real-time security integration, reduced manual intervention, and increased operational resilience in DevSecOps pipelines.”
From Evolution of DevSecOps and Its Influence on Application Security: A Systematic Literature Review (2025) · cites this paper as [62]
The evaluation is largely exploratory and does not provide extensive empirical validation or benchmarking against other AI-driven security tools.
“The evaluation is largely exploratory and does not provide extensive empirical validation or benchmarking against other AI-driven security tools.”
From Evolution of DevSecOps and Its Influence on Application Security: A Systematic Literature Review (2025) · cites this paper as [62]
Cites
LogFiT: Log Anomaly Detection Using Fine-Tuned Language Models
Almodovar, C., Sabrina, F., Karimi, S., & Azad, S. (2024) · IEEE Transactions on Network and Service Management · doi:10.1109/tnsm.2024.3358730 · Cited by 111 · see citing works
GPT-4 vs. GPT-3Preprint
Koubâa, A. (2023) · doi:10.36227/techrxiv.22312330.v2 · Cited by 80 · see citing works
Embracing IaC Through the DevSecOps Philosophy: Concepts, Challenges, and a Reference Framework
Alonso, J., Piliszek, R., & Cankar, M. (2023) · IEEE Software · doi:10.1109/ms.2022.3212194 · Cited by 30 · see citing works
Security in DevSecOps: Applying Tools and Machine Learning to Verification and Monitoring Steps
Cankar, M., Petrovic, N., Pita Costa, J., Cernivec, A., Antic, J., Martincic, T., & Stepec, D. (2023) · Companion of the 2023 ACM/SPEC International Conference on Performance Engineering · doi:10.1145/3578245.3584943 · Cited by 24 · see citing works
LogLG: Weakly Supervised Log Anomaly Detection via Log-Event Graph Construction
Guo, H., Guo, Y., Yang, J., Liu, J., Li, Z., Zheng, T., Zheng, L., Hou, W., & Zhang, B. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-30678-5_36 · Cited by 22 · see citing works
Chat GPT-Based Design-Time DevSecOps
Petrović, N. (2023) · International Scientific Conference on Information, Communication and Energy Systems and Technologies (ICEST) · doi:10.1109/icest58410.2023.10187247 · Cited by 17 · see citing works
Self-citationAutomated Approach to IaC Code Inspection Using Python-Based DevSecOps Tool
Petrović, N., Cankar, M., & Luzar, A. (2022) · Telecommunications Forum (TELFOR) · doi:10.1109/telfor56187.2022.9983681 · Cited by 15 · see citing works
Self-citationAutomated Log Analysis and Anomaly Detection Using Machine Learning
Shah, A. H., Pasha, D., Zadeh, E. H., et al. (2022) · Frontiers in artificial intelligence and applications · doi:10.3233/faia220378 · Cited by 12 · see citing works
LogQA: Question Answering in Unstructured LogsPreprint
Huang, S., Liu, Y., Fung, C., Qi, J., Yang, H., & Luan, Z. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2303.11715 · Cited by 8 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Generative Artificial Intelligence for Software Engineering—A Research Agenda
Nguyen‐Duc, A., Cabrero‐Daniel, B., Przybylek, A., Arora, C., Khanna, D., Herda, T., Rafiq, U., Melegati, J., Guerra, E., Kemell, K., et al. (2025) · Software: Practice and Experience · doi:10.1002/spe.70005 · Cited by 60 · see citing works
Evolution of DevSecOps and Its Influence on Application Security: A Systematic Literature Review
Mohammed, K., Shanmugam, B., & El-Den, J. (2025) · Technologies · doi:10.3390/technologies13120548 · Cited by 8 · see citing works
- Authors
- Petrović, N.
- Affiliations
- University of Nis (RS)
- Corresponding author
- Petrović, Nenad
- ORCID
- Petrović, Nenad
- Venue
- 2023 10th International Conference on Electrical, Electronic and Computing Engineering (IcETRAN)
- Publisher
- IEEE
- Published
- 2023-06-05
- Pages
- 1-5
- OpenAlex
- W4385323797
- Cited by
- 17
- References
- 9
- Citations by year
- 2026: 1 · 2025: 7 · 2024: 4 · 2023: 5
- FWCI
- 5.14
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- International Conference on Electrical, Electronic and Computing Engineering (IcETRAN) (publishedVersion)
- Funders
- Ministry of Education
- Topics
- Physical Sciences › Computer Science › Computer Networks and Communications › Network Security and Intrusion Detection
- Physical Sciences › Computer Science › Computer Networks and Communications › Software System Performance and Reliability
- Physical Sciences › Computer Science › Artificial Intelligence › Adversarial Robustness in Machine Learning
- Keywords
- DevSecOps, server log analysis, suspicious activity detection, ChatGPT, large language models, intrusion detection, log classification
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
