GENERATIVE AI AND PHD SUPERVISION – PERSONALISATION AND ALIENATION, BENEFITS AND RISKS

Doctoral supervision is a primary factor when analysing the doctoral process completion and success. We expect more and more from doctoral graduates – good research, societal impact, policy advocacy and above all multidimensional, interdisciplinary and highly adaptive specialists. All these require an excellent supervision. In time, the doctoral mentoring evolved from having a unique supervisor to multiple possible structures – one supervisor and a guiding committee, two supervisors – cotutelles, more than two supervisors in joint programs with several institutions, one academic supervisor and one industry supervisor, etc. Lately, a very “special” PhD mentor joined the doctoral team – Generative Artificial Intelligence (AI). Friend or foe, for the doctoral student or the supervisor, or for both (the team), AI tools and models become more and more the new normal in doctoral research. Is this good for PhD supervision, can it add to the so much necessary personalisation of doctoral supervision, or is it just another factor contributing to the lack of real socialisation, leading to supplementary alienation? The present essay is an attempt to analyse the benefits and risks of using AI in doctoral supervision, based on the extant literature – still very scarce, due to the novelty of the subject. Our results will be presented based on the structure of a classical awareness-understanding matrix, with three sections – what we know, known unknowns and unknown unknowns. Findings could help reflect upon and design better policies in the field, answering to ethical dilemmas, to support and protect, at the same time, both doctoral researchers and their supervisors.

DOI: https://doi.org/10.47743/rebs-2025-2-0005