Beyond the arms race: An exploratory study on shadow AI usage and competency risks among DBA students in Sarawak

Abstract

The advent of generative artificial intelligence (AI) is affecting postgraduate education, particularly professional doctorates, which demand that candidates juggle full-time employment with the rigors of academic study. Generative artificial intelligence can improve efficiency and facilitate analytical thinking but also presents challenges regarding academic integrity, supervision, and the quality of doctoral learning. This study is a qualitative exploratory study exploring the phenomenon of shadow AI, the informal and covert use of generative AI for academic work, among Doctor of Business Administration students in Sarawak, Malaysia. Semi-structured interviews were conducted with five working DBA students who actively used generative AI tools. Data were analyzed using the constant comparative method, and four themes emerged: pragmatic survival, the verification shield, competency risk, and systemic disconnect. The findings suggest that students’ engagement with shadow AI was not mainly intended to facilitate cheating but rather served as a means of dealing with workload pressures, reconciling professional and student identities, and managing ambiguous institutional policies on AI and academic integrity. However, long-term hidden dependence can erode cognitive capacity, professional judgment, and doctoral-level expertise. Focusing on detection may push AI further underground. The study synthesizes Shadow AI and Conservation of Resources theories to advocate transparent, human-centered, and assessment-based policies that acknowledge generative AI as a fixture of doctoral education while protecting independent thought, ethical practice, academic rigor, and core doctoral competencies.