“It’s Not Just a Technology, but a Team Member”: Generative AI Use Among Students in CPED Doctor of Education Programs

The rapid adoption of generative artificial intelligence (genAI) in higher education has created both opportunities and challenges for Doctor of Education (EdD) students and their institutions. While EdD students increasingly integrate genAI tools like ChatGPT into their academic work, their usage patterns, ethical decision-making processes, and critical evaluation strategies remain poorly understood. This study explored how EdD students in Carnegie Project on the Education Doctorate (CPED) institutions use genAI and how those practices relate to institutional guidance and faculty-student relationships.Guided by a conceptual framework of metaliteracy and generative AI literacy, this study explored genAI use across cognitive, behavioral, affective, and metacognitive domains. Survey data were collected from 190 EdD students across North American CPED institutions, followed by four focus groups with 20 participants. The survey measured usage patterns, ethical awareness, and evaluation strategies, while focus groups provided deeper insight into decision-making and help-seeking behaviors.
Findings revealed that expectations for generative AI use and acknowledgment were often unclear across institutions and programs. In the absence of specific and consistent guidance, students developed their own personal guidelines for use, grounded in scholarly values such as originality, expertise, and voice. Participants described using generative AI not only as a tool for specific tasks, but as 24/7 academic support. Many reported turning to it first for feedback, clarification, and idea development before reaching out to faculty or peers, though they described varied approaches to transparency, often shaped by concern about how their use might be perceived. Although students expressed confidence in their ability to evaluate AI-generated output, their descriptions revealed uneven technical understanding.
These findings informed the Shifting Pedagogical Interactions and Relational Instructional Ties (SPIRIT) Model that illustrates how institutional ambiguity and identity-driven decision-making contribute to the integration of genAI as frontline academic support. As students increasingly use genAI for tasks traditionally associated with faculty mentoring and guidance, the sequencing of faculty-student interaction shifts. Rather than replacing faculty mentorship, these patterns suggest a reorganization of academic support pathways within CPED-aligned doctoral programs. The study concludes with implications for doctoral program design and future research in AI-mediated higher education.