Recently I hosted an Academic Writing and Knowledge Exchange Network (AcWriKEN) session on GenAI in academic writing provision for researchers. I came in with slides on frameworks, policy gaps and what we’re doing at GCU. I left thinking much more about feelings than frameworks. Which, if you know me, isn’t entirely surprising.
We ran the session under the Chatham House Rule, so what follows is what I heard, not who said it. Thank you to everyone who joined and shared so openly.
It’s muddled out there
The first thing that struck me was just how muddled this space is. Across the breakout rooms, most of us don’t yet have specific or embedded GenAI provision for researchers. Some institutions have brought in external sessions, others are handling it case by case, and there’s a growing expectation that we “weave it in” somewhere. Several people said, quite honestly, that they’re only just starting to explore what that might look like.
The policy picture isn’t much clearer. People draw on institutional guidance for day-to-day questions, publishers for guidance on outputs and funders to understand what’s expected. But institutional guidance keeps changing, so you have to actively follow it, and guidance for postgraduate researchers (PGRs) tends to arrive more slowly than everything else. One question came up that I’ve not stopped thinking about: what actually is “the standard”?
At GCU we’ve noticed that PGRs often sit between two documents. Staff guidance is written for staff, assessment guidance is written for taught students, and a doctoral thesis doesn’t fit neatly into either. From the discussion today, we’re not alone in that. Policy tends to be generalised, while PGR needs, uses and circumstances are much more nuanced. Sustainability came up too, as a concern that rarely makes it into guidance at all.
The bit we forget
Here’s what really stayed with me. We spend a lot of time on what’s allowed and how to declare it. We spend far less time on what it actually feels like to be a researcher navigating all of this right now.
And the feelings that came through were shame, stress and worry.
Shame about using AI at all, or about not being sure whether what you did “counts”. Stress about getting it wrong when the rules are unclear and keep shifting, and when the consequences might be an integrity conversation. Worryabout being left behind if you don’t use it, and about losing your own voice or skills if you do. In my slides I called this the tension between FOMO and integrity, “Am I being left behind?” versus “What am I actually responsible for?”. Hearing it in the rooms made it feel a lot less abstract.
Then there’s the “forced” use. GenAI isn’t something researchers always choose to pick up anymore. It’s switched on by default in the tools they already use for writing, email and searching. It’s in the search results, the reference manager, the word processor. So a researcher can end up “using AI” without ever making a conscious decision to, and then feel anxious about whether they need to declare it. That’s a very different starting point from the tidy “decide, then use” model our guidance often assumes.
I think we underestimate how much of this researchers carry quietly into our workshops.
What this means for our practice
So how do we better build this into how we deliver workshops and develop guidance? Here’s where my thinking has got to so far, and I’d genuinely love to hear yours.
Name the feelings out loud. A quick moment at the start of a session acknowledging that this is confusing, that guidance is still catching up and that uncertainty is normal can lower the temperature a lot. People are much more likely to ask the question they’re actually worried about if they don’t feel they’re the only one who doesn’t know.
Make not using AI a legitimate choice. In the AIR framework from Electv Training, A0 (no AI) is a band in its own right, and higher bands are not better. I’d like us to say that more often. Choosing not to use AI for a task, because you want to build the skill or because you have concerns about sustainability or data, is a researcher making a considered decision, not someone being left behind.
Give people permission to pause. In our TASK framework, one question giving you pause is enough reason to stop and reconsider. I’ve started to see that line as a wellbeing message as much as an integrity one. Pausing isn’t failing.
Help people notice the defaults. If AI is built into the tools, then part of our provision is helping researchers spot where it’s already operating and decide what they want to do about it. That moves them from “it just happened” back to “I decided”, which is where the sense of control (and much of the confidence) comes from.
Expect disagreement, and say so. When we run the same scenario in our sessions, different people reach different verdicts, and disciplines diverge quickly. If we present that as normal and useful, rather than as someone being “wrong”, it takes some of the shame out of the room.
Keep rules and development together. Most PGR guidance across the sector is framed around assessment: what’s permitted, record keeping, declarations. There’s very little on development, on reading, note-making, critical thinking and writing. That developmental gap is exactly where academic writing provision can contribute, and it’s also where the confidence gets built.
Bring supervisors into the same conversation. A lot of worry comes from not knowing what your supervisor thinks. Shared language, whether that’s TASK, AIR or something else, means both sides of the supervisory conversation can talk about AI use without it feeling like a confession.
Write guidance with the reader’s state of mind in mind. When we draft guidance, it’s worth asking how it lands for an anxious first-year PGR rather than just whether it’s accurate. Plain language, worked examples and a clear “if you’re unsure, here’s who to ask” go a long way.
Where next
Which leaves me with a question for you: how do you make room for the emotional side of GenAI in your own workshops and guidance? And if you’re a PGR or researcher reading this, what would actually help you feel less worried?
Join the AcWriKEN Group on LinkedIn or sign up to our AcWriKEN Jiscmail.
