Strategies for Managing Technical Debt in the Scaling of Automated Information Processing Systems Using Artificial Intelligence Methods
Keywords:
technical debt, artificial intelligence, scalable systems, data, architecture, causal modeling, system managementAbstract
The article presents an analysis of contemporary approaches to managing technical debt in the scaling of automated information processing systems using artificial intelligence methods. The study is conducted in the format of a systematic review and analytical synthesis of scientific publications addressing technical debt in AI systems, scalability issues, and the management of data, models, and architecture. Particular attention is given to the interrelationship between the stages of technical debt formation, system structure, and the quality of decision-making. The main sources of debt accumulation are examined, including errors at the requirements level, data degradation, model instability, increasing architectural complexity, and methodological limitations associated with substituting causal relationships with correlations. It is established that under scaling conditions, technical debt does not lead to system failures but manifests through behavioral distortions and a decline in decision quality. It is substantiated that management effectiveness is determined not by the elimination of individual defects, but by the coordination of actions across all system levels and the ability to constrain the propagation of debt. An original multi-level model of technical debt management is proposed, reflecting the processes of its formation, diagnosis, management, and prevention, while accounting for its impact on system behavior. The model demonstrates a transition from reactive mitigation to proactive architectural management based on the integration of MLOps, data governance, enhanced model transparency, and the use of causal modeling.
References
[1]. Alassuli, A., Eltweri, A., Thuneibat, N. S., Al-Hajaya, K., & Ismail, S. M. (2026). Artificial intelligence applications and financial forecasting accuracy in banking platforms: Evidence from Jordan. Administrative Sciences, 16(3), 122. https://doi.org/10.3390/admsci16030122
[2]. Alenazi, M. (2025). The role of environmental assumptions in shaping requirements technical debt. Applied Sciences, 15(14), 8028. https://doi.org/10.3390/app15148028
[3]. Bogner, J., Verdecchia, R., & Gerostathopoulos, I. (2021). Characterizing technical debt and antipatterns in AI-based systems: A systematic mapping study. arXiv. https://doi.org/10.48550/arXiv.2103.09783
[4]. Borowa, K., Ratkowski, A., & Verdecchia, R. (2025). The technical debt gamble: A case study on technical debt in a large-scale industrial microservice architecture. arXiv. https://doi.org/10.48550/arXiv.2506.16214
[5]. de Oliveira, N. A., & Basso, L. F. C. (2025). Advancing credit rating prediction: The role of machine learning in corporate credit rating assessment. Risks, 13(6), 116. https://doi.org/10.3390/risks13060116
[6]. Ernst, N. A., Awon, A. M., Hingmire, S., & Li, Z. S. (2026). The nature of technical debt in research software. arXiv. https://doi.org/10.48550/arXiv.2603.20415
[7]. Gordieiev, O., Gordieieva, D., Rainer, A., Gorbenko, A., & Tarasyuk, O. (2026). Quality assessment of artificial intelligence systems: A metric-based approach. Electronics, 15(3), 691. https://doi.org/10.3390/electronics15030691
[8]. Holzinger, A., Longo, L., Cangelosi, A., & Ser, J. D. (2026). Research frontiers in machine learning & knowledge extraction. Machine Learning and Knowledge Extraction, 8(1), 6. https://doi.org/10.3390/make8010006
[9]. Huh, J. (2026). Prescriptive analytics for sustainable financial systems: A causal–machine learning framework for credit risk management and targeted marketing. Systems, 14(1), 16. https://doi.org/10.3390/systems14010016
[10]. Moreschini, S., Arvanitou, E. M., Kanidou, E.-P., Nikolaidis, N., et al. (2025). The evolution of technical debt from DevOps to generative AI: A multivocal literature review. Journal of Systems and Software, 231, 112599. https://doi.org/10.1016/j.jss.2025.112599
[11]. Recupito, G., Pecorelli, F., Catolino, G., Lenarduzzi, V., Taibi, D., Di Nucci, D., & Palomba, F. (2024). Technical debt in AI-enabled systems: On the prevalence, severity, impact, and management strategies for code and architecture. Journal of Systems and Software, 219, 112151. https://doi.org/10.1016/j.jss.2024.112151
[12]. Shivashankar, K., & Martini, A. (2026). TD-Suite: All batteries included framework for technical debt classification. In A. Goldman et al. (Eds.), Agile processes in software engineering and extreme programming (XP 2026) (Lecture Notes in Business Information Processing, Vol. 578). Springer. https://doi.org/10.1007/978-3-032-22375-3_9
[13]. Simon, E. I. O., Hettiarachchi, C., Potanin, A., et al. (2026). Automated detection of algorithm debt in deep learning frameworks: An empirical study. Empirical Software Engineering, 31, 66. https://doi.org/10.1007/s10664-026-10807-5
[14]. Sklavenitis, D., & Kalles, D. (2024). Measuring technical debt in AI-based competition platforms. In Proceedings of the 13th Hellenic Conference on Artificial Intelligence (SETN ’24) (Article 27, pp. 1–10). https://doi.org/10.1145/3688671.3688783
[15]. Sklavenitis, D., & Kalles, D. (2025). A scoping review and assessment framework for technical debt in the development and operation of AI/ML competition platforms. Applied Sciences, 15(13), 7165. https://doi.org/10.3390/app15137165
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Dmitrii Bezfamilnyi
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Authors who submit papers with this journal agree to the following terms.
