Mitigation of Artificial Intelligence Technical Debt and Its Impact on the Success of Institutional Information Systems During the Digital Transformation of Higher Education
Year: 2026
Type: conference-paper
Abstract: Higher Education Institutions (HEIs) increasingly adopt Large Language Models (LLMs) to modernize legacy IT infrastructures. While LLMs accelerate Digital Transformation (DT), they introduce hidden AI technical debt (AI-TD), including data, architectural, and process debt, which threatens longterm Information Systems (IS) success. This study aims to quantify how Software Engineering (SE) principles mitigate this emerging debt and influence overall IS effectiveness. Integrating the DeLone and McLean IS Success Model with Enterprise Architecture (EA) paradigms, this study adopts an explanatory sequential mixed-methods approach. In the initial quantitative stage, Partial Least Squares Structural Equation Modeling (PLS-SEM) is used to assess the causal relationships between SE principles, AI-TD, and IS success dimensions across HEIs. A subsequent qualitative phase uses thematic analysis of expert interviews to contextualize these statistical findings. Preliminary structural modeling suggests that rigorous SE principles are expected to mitigate AI-TD and architectural degradation, while high EA maturity is expected to moderate the negative impacts on system and information quality. Ultimately, this research aims to provide a novel theoretical extension to IS success metrics and equip IT leaders with a validated framework for sustainable, agile, and strategically aligned institutional modernization.
Authors: Alejandro Aldas-Alarcon, Iván Carrera
Institution: National Polytechnic School
Language: English
Subfield: Management Information Systems
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W7214727862