Postgraduate Education Practitioners Network Clinic — January 2026

The next in our (roughly) quarterly series of PEP Network clinics – which aim to provide a safe and supportive environment for members to connect with one another and to share questions, concerns, and insights – saw more than 40 colleagues from across the sector come together to discuss their experiences of responding to the proliferation in the use generative artificial intelligence (GenAI) by postgraduate researcher (PGRs). The session was structured around four broad questions:

• Has your institution developed policy around the use of GenAI, targeted at PGRs, supervisors, and/​or examiners?

  Responses in the virtual room indicated a lean towards institutional position statements and/​or guidance (rather than policy per se) in relation to PGRs, with this documentation often sitting alongside updates to more generalised all-student policies around academic integrity, academic misconduct, etc. This was for a mix of reasons, including the need to remain responsive in a fast-moving space, challenges in generating momentum in institutional contexts, and the potential difficulty in proving the use of GenAI undermining the enforceability of policies.

• Has your institution developed policy around the use of GenAI in PGR applications?

  Responses were varied in relation to the use of GenAI in PGR applications. Any policy updates referenced were generally short statements indicating that GenAI should not be used, but participants noted that this may be difficult to pick up in practice, while the intersection with potential legitimate use (e.g. accessibility) was also raised. With this in mind, the use of declarations in applications to improve transparency was discussed.

  When considering other forms of mitigation, one institution reported a possible move to requiring applicant interviews more often, while another noted experience of witnessing AI use (via double-screening) in an interview context and the potential for this to undermine the interview process. The looming challenge of AI avatars was also raised, both in the context of interviews and – later on down the line – vivas. While awareness is building around this possibility, colleagues acknowledged that it is not one they have addressed in guidance, policy, or training as yet.

• Has your institution developed training for PGRs around GenAI?

  A number of institutions indicated that training is in development, with ownership of that training varying between PGR-focussed, library and education teams. The benefit of in-person training was considered against the need to achieve reach. One institution noted that they hold regular AI forums for PGRs to gather and learn together through sharing practice, asking questions, etc.

  The conversation then moved in to a discussion about staff training, with colleagues acknowledging their own lack of knowledge and/​or confidence in such a fast-moving space, and the knock-on effect this can have on the development of policy, guidance and/​or training.

• Has your institution experienced challenges to open access and/​or embargo policies in relation to GenAI?

  Two institutions reported instances of PGRs pushing back on having their theses available on institutional repositories due to concern about data scraping. The balance between open access obligations and legitimate concerns about large language models was briefly discussed. It was also acknowledged that increasing PGR training in GenAI may lead to more instances of pushback about the availability of research outputs, and so other institutions indicated that they would take this away for consideration in their own contexts.

Thank you to everyone who attended today; we are likely to return to this topic again in the future. Following a suggestion at a previous clinic, we will focus on international student support at the next session in April.

We do want these sessions to be driven by your needs, so please feel free to suggest topics for future clinics using the form linked below.