Abstract:

The rapid emergence of generative Artificial Intelligence (AI) since 2022 has introduced profound changes in higher education, particularly in how students and academic staff engage in teaching, learning, and knowledge production. While numerous studies have examined the impact of AI on academic writing, assessment, and feedback automation, the specific domain of academic supervision—especially the supervision of undergraduate and master’s theses—remains significantly underexplored. This study presents the protocol and preliminary findings of a scoping review aimed at mapping current evidence on the use of generative AI in thesis supervision, mentoring, and academic support processes in university settings.

The review follows the PRISMA-ScR guidelines to ensure methodological transparency and adopts the PCC framework (Population–Concept–Context) to define eligibility criteria:
(P) students and supervisors in higher education,
(C) use of AI-based or AI-assisted tools in supervision and feedback processes, and
(C) university-level academic contexts.

Searches were conducted in Web of Science focusing on publications from 2022 to 2025 in English and Spanish, coinciding with the period in which generative AI tools became widely accessible to the academic community.

Preliminary results reveal a rapidly evolving landscape in which AI is being used for a wide range of supervisory practices. These include generative chatbots providing methodological guidance, intelligent tutoring systems assisting with structuring research questions, and analytic platforms supporting iterative feedback and writing improvement. Reported benefits include increased personalization, enhanced accessibility for students requiring frequent support, and more efficient feedback cycles. However, the literature also highlights substantial challenges, particularly concerning authorship ethics, academic integrity, transparency, and the redefinition of the supervisor’s role in an AI-mediated environment.

These initial findings contribute to the development of AI literacy initiatives for tutors and students and offer a conceptual foundation for ethical and pedagogically informed supervision models. The study also identifies critical gaps in current research, underscoring the need for empirical analyses of supervisory relationships, power dynamics, and new forms of academic dependency.

Keywords:

Artificial intelligence, higher education, academic supervision, thesis tutoring, academic integrity, educational innovation.