As AI tools become more embedded in research practice, we’re rightly asking: Can they help with summarising? Coding? Theme development? Interpretation?
But there’s a quieter shift happening too — one we’re not talking about enough.
As tools like ChatGPT, Claude, and MAXQDA’s AI Assistant become standard parts of the workflow, they’re not just supporting the work. They have the potential to reshape what we value as good qualitative research — what counts as rigour, what counts as insight, even what we think “analysis” is.
This blog is about noticing the epistemic drift that can happen when we don’t pause to ask how the tools are shaping the field itself.
When Fast Starts Looking Like Rigour
AI is extremely good at identifying surface-level patterns. It clusters keywords, suggests themes, even writes summaries that sound impressively polished.
That can be genuinely useful — especially at the start of a project or when working with large datasets.
But the risk is this: the more we see those outputs, the more they start to feel like analysis. Over time, we can begin to mistake pattern recognition for interpretation, tidy themes for depth, and fluency for credibility.
The question then becomes: what role is AI actually playing in your analysis?
This doesn’t happen suddenly. But when AI fits a certain model — descriptive, repeatable, tidy — and we start building our expectations around that model, we risk skewing what we consider “rigorous” without realising it.
Not Just About Tools — About Values
What we reward, expect, or replicate becomes part of our collective method. If reviewers or supervisors start valuing AI-enhanced efficiency, it can quietly displace practices that take longer but offer more depth — like exploring research paradigms, reflexive journelling, or sitting with discomfort.
This isn’t about gatekeeping. It’s about remembering that rigour in qualitative research isn’t speed, or neatness. It’s about reflexivity, choosing the right approach to answer the research question, and depth of engagement with the data.
Reflexive Use Means Staying Alert to the Frame
AI is not neutral. It brings assumptions, about what counts as knowledge, about how language either reflects or creates meaning, about what kind of analysis is “good enough.”
To use it well, we need to:
- Lead the analysis, not the other way around
- Choose to retain the beautiful inefficiencies — the looping, memoing, and pausing — that actually make sense-making possible
AI can help. It can accelerate, support, scaffold. But it can’t decide what matters. As qualitative researchers, we bring our subjectivity, and reflexivity to our research, and the responsibility for what gets counted as knowledge.
Final Thought: Stay Grounded in the Why, Not Just the How
If AI is part of your research process, fantastic — just stay grounded in your own analytic values. Use the speed and fluency to your advantage, but don’t let them become the standard of quality.
The most powerful work still comes from your engagement with the data, reflexivity, and epistemic humility.
🟡 Curious about how to integrate AI tools into your qualitative work — without losing depth or rigour? The AI + Qualitative Research Masterclass is designed for exactly that.
→ Register now (Australia)
