Notes from my webinar: Qualitative Research in 2026: New Challenges & Opportunities

Photo by Yen Vu on Unsplash
I began my webinar earlier in October by explaining what the webinar wouldn’t do in our discussion of qualitative research and GenAI. It wouldn’t promote GenAI in qualitative research, or offer a toolkit, a workflow or quick ‘short cuts’. Qualitative research is not a chore to be sped up or automated.
Instead, I wanted to create a supportive space to think about where qualitative research is today, the threats and risks to it (including from GenAI), how we want to shape the field, and how we support newcomers. I wanted us all to pause and consider:
“What kind of qualitative researcher do I want to be?”
The ‘Columbus effect’
You may have come across the ‘Columbus generation’, a nickname for Gen Z’s habit of discovering things that already exist. Urban Dictionary’s example is “Quiet Walks”, where you go for a walk without your technology, otherwise known as just walking (O’Connor 2026).
In some ways, we are made to believe that qualitative research is having its own Columbus moment. The GenAI tech industry is rediscovering and reinventing qual, selling solutions to problems that don’t exist, and often misunderstanding the principles of qual in its claims.
Much of this rests on myths that come from assuming quantitative principles are how all research is done. These include the ideas that more is better and research should happen ‘at scale’, that we should strive for neutrality and objectivity and strip out bias, that methods should be standardised, and that you don’t need a human to interview or facilitate. Most qual is none of these things. It’s not a task or chore to be sped up or automated.
Myth busting: what qualitative research is
Qualitative research is a diverse field, with many different methodologies and methods. It focuses on understanding meaning and meaning-making, and it thrives when exploring experience and difference. It isn’t about measurement, counting or frequencies. It’s reflexive and iterative, and it’s creative and contextual. Often it involves researching with participants through co-creation and participatory methods. And it’s human, relying on empathy, rapport, trust and ethics.
That’s why we shouldn’t compare apples with oranges. A qualitative study can’t be judged against positivist, quantitative criteria. We evaluate it on the methodology and methods the researcher chose to answer a specific question in a particular context.
No research is neutral, either. Western research has a long history as a tool of imperialism and violence. For example:
Linda Tuhiwai Smith (1999: 1) described ‘research’ as “probably one of the dirtiest words in the indigenous world’s vocabulary”.
Feminist researchers such as Ann Oakley (1981), and Liz Stanley and Sue Wise (1993), challenged the myth of ‘hygienic research’ decades ago. As I wrote in my Reflexivity book:
“By being reflexive we acknowledge that we cannot be separated from our biographies” (Lumsden 2019: 23).

Photo by Royynne O on Unsplash
The harms and risks of GenAI
In the second part of the webinar we explored the harms and risks of GenAI:
1. Environmental harms
Data centres used around 415 TWh of electricity in 2024, about 1.5% of global demand, and this is projected to more than double by 2030, driven largely by AI (IEA 2025). Cooling them takes vast amounts of fresh water, often in water-stressed regions (Li et al. 2023). AI also depends on mined minerals and generates growing electronic waste (Crawford 2021).
2. Labour exploitation
AI relies on hidden ‘ghost work’ by low-paid workers who label data and train models (Gray and Suri 2019). Kenyan workers filtering violent and abusive content to make ChatGPT ‘safer’ were paid less than $2 an hour (Perrigo 2023). This labour is concentrated in the Global South, reproducing colonial patterns of extraction (Muldoon, Graham and Cant 2024).
3. Harms to the research
GenAI flattens the richness of human experience. It risks loss of meaning, misinterpretation and a loss of reflexivity. It’s also worth asking whether it really saves time and money once you’ve checked everything it produces. Moreover, these tools are ‘black boxes’ that can hallucinate and misread what participants meant. They also can’t interpret latent meanings, pauses, silences, or absences.
4. Harms to participants
Consent processes don’t always account for GenAI tools retaining data for training. Do you know what your participants think about GenAI? Do you have permission to input their i.e. interview transcripts into GenAI tools? Can you really trust that this means the data won’t be used for training (now or in the future)? Without transparency and honesty, we risk harming communities and the trust we built with them.
5. Harms to knowledge and power
Whose voices, experiences and knowledge are legitimised, reproduced or erased when GenAI is developed and used? These tools are built on human biases and errors, and on Western knowledge.
6. Slopification and homogenization
Sourati et al. (2026) describe GenAI’s homogenising effect on human expression and thought as a form of ‘McDonaldization’. In qual, this looks like quick, churned-out ‘qualitative’ research ‘at scale’, based on misunderstandings of what qual is and how we do qual.
7. Collective amnesia and deskilling
What creative new approaches to qual might a reliance on AI distract us from? And how well do we know our own research when GenAI has done part of it for us? Laurel Richardson (1994: 923) reminds us that writing “is not just a mopping-up activity at the end of a research project”. It’s a way of knowing and a method of discovery, and Joanne Yoo (2025) shows how we ‘become’ through our writing craft.
Resurfacing: what this means in practice
My stance is for human-led qualitative research. That means avoiding GenAI where possible and being aware of its harms. We each need to make deliberate, informed choices rather than being swept along. I suggested some tips for navigating these choppy waters:
- Don’t get caught up in FOMO: GenAI sales pitches are designed to make you feel left behind.
- Check the AI policies of your organization, funder, publishers, etc… and keep checking them! Look at guidance from your organisation, your university and ethics committee, academic publishers, and professional associations.
- Ask yourself two questions: Can I do this without AI? And do I have permission to use it?
- Keep an audit trail: If and when you do use AI, record when, where and why.
- Remember that AI doesn’t make your research ‘better’ or ‘more rigorous’: For example, many qual researchers don’t turn to software and/or AI for their qualitative analysis – find the best approach for you.
- Learn to do it properly first: No tool can replace your understanding of your own work. PhD examiners, funders and reviewers will still ask why you chose that method or theory, what the limitations were, and what you’d do differently. You need to be able to justify and defend your work and decisions.
- Return to your values: For me these include transparency and honesty, recognising the footprints we leave in our work and the qualitative field, and a focus on co-creation. Put reflexivity into practice, and extend it to how and why you do or don’t use GenAI.
The words of Sherry Turkle resonated with me in an interview about her new book, Artificial Intimacy: Who We Become When We Talk to Machines (2026), which explores human relationships with technology. Her advice was to keep AI “in its lane”, and to reflect critically on when we use it at all. And if you are tempted to take it out of its lane, she suggests thinking about what a human could do for you instead.
As Emily Bender and Alex Hanna (2025) acknowledge, pushing back and resisting the AI hype can feel lonely. But they also remind us that resisting hype can be “empowering, grounding, and even joyful” (p. 196). It reaffirms the value of our skills, and of being human together.
From the discussion: can GenAI ever not be extractive?
One of the most pertinent points raised in the discussion was about extraction. Qualitative research has long pushed back against extractive ways of doing research, recentering the human through decolonial and feminist ethics of care. But now we’re now being pushed towards tools that are, at their core, extractive.
Consider: whose data, writing and creative work gets scraped to train these tools, often without consent, and who profits from it. These tools are also built on human bias, in how they’re designed, what goes into them, and in who gets to make those decisions. Even at the everyday level, using them involves a choice about our values as researchers. GenAI creep sees these tools integrated into the everyday tech we use, and often with us now having to ‘opt-out’ (i.e. Google searches), thus normalising their presence in our tech use.
One webinar participant raised a related worry. If choosing not to use GenAI is a matter of conscience (because of its climate impact, extraction, exploitative labour practices and links to military uses), how will researchers who opt out be seen? Will they be judged as behind the times, or as less employable? And do we need protective policies so that researchers can say no? Others in the chat described employers already pushing them to use AI, and some academic jobs pivoting towards it.
But there are spaces of resistance worth seeking out. We have the work of sociologist Deborah Lupton. And Janet Salmons – who is reviewing academic publishers’ AI and data privacy policies. Others pointed to the DAIR Institute and to Alex Hanna’s work on data workers, and suggested we start thinking of ourselves as research workers who can organise too. In 2019, 419 researchers signed the statement published in Qualitative Inquiry which rejects the use of GenAI in qualitative work (see Jowsey et al. 2025).
Here at Qualitative Training, we want to embrace and provide spaces where it’s normal to ask these questions, and where nobody is left feeling like the only one in the room if they decide not to use GenAI.

Photo by Eric Prouzet on Unsplash
Questions to take away
I closed the webinar with these questions, and I’ll leave you with them too:
- What kind of qualitative researcher do you want to be?
- What future do we want to shape as a community of qualitative researchers?
- What footprints do you leave (on the field)?
I’d love to read your thoughts on how you feel about AI in qualitative research. And also how to carefully consider its use or non-use. Comment below. (Note: any comments promoting or selling AI tools will be deleted).
Next steps
My next free webinar is “Ask a human, not a bot”, a Qualitative Training webinar on Reflexive Thematic Analysis, on Wednesday 11 November at 13.30 (GMT). Register here and/or sign up to our newsletter to find out more about future events.
And if you’d like support with your PhD, professional research or viva prep, get in touch here. Go to a real human for support, not a bot!
References
Bender, E.M. & Hanna, A. (2025) The AI Con: How to Fight Big Tech’s Hype and Create the Future We Want. London: The Bodley Head.
Crawford, K. (2021) Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. New Haven and London: Yale University Press.
Gray, M.L. & Suri, S. (2019) Ghost Work: How to Stop Silicon Valley from Building a New Global Underclass. Boston: Houghton Mifflin Harcourt.
International Energy Agency (2025) Energy and AI. Paris: IEA. URL (accessed 28 October 2026: https://www.iea.org/reports/energy-and-ai
Jowsey, T., Braun, V., Clarke, V., Lupton, D., & Fine, M. (2025). ‘We Reject the Use of Generative Artificial Intelligence for Reflexive Qualitative Research.’ Qualitative Inquiry. https://doi.org/10.1177/10778004251401851
Li, P., Yang, J., Islam, M.A. & Ren, S. (2023) ‘Making AI Less “Thirsty”: Uncovering and Addressing the Secret Water Footprint of AI Models’. arXiv preprint: https://doi.org/10.48550/arXiv.2304.03271
Lumsden, K. (2019) Reflexivity: Theory, Method, and Practice. London: Routledge.
Muldoon, J., Graham, M. & Cant, C. (2024) Feeding the Machine: The Hidden Human Labour Powering AI. Edinburgh: Canongate.
Oakley, A. (1981) ‘Interviewing Women: A Contradiction in Terms?’ In H. Roberts (Ed.) Doing Feminist Research (pp. 30–61). London: Routledge and Kegan Paul.
O’Connor, R. (2026) ‘How Gen Z Became Known as the “Christopher Columbus Generation”‘. Newsweek, 2 May 2026. URL (accessed 28 October 2026): https://www.newsweek.com/how-gen-z-became-known-as-the-christopher-columbus-generation-11897766
Perrigo, B. (2023) ‘Exclusive: OpenAI Used Kenyan Workers on Less Than $2 Per Hour to Make ChatGPT Less Toxic’. TIME, 18 January. URL (accessed 28 October 2026): https://time.com/6247678/openai-chatgpt-kenya-workers/
Richardson, L. (1994) ‘Writing: A Method of Inquiry’. In N.K. Denzin and Y.S. Lincoln (Eds.) Handbook of Qualitative Research (pp. 516–529). Thousand Oaks, CA: Sage.
Smith, L.T. (1999) Decolonizing Methodologies: Research and Indigenous Peoples. London: Zed Books.
Sourati, Z.S., Ziabari, A. & Dehghani, M. (2026) ‘The Homogenizing Effect of Large Language Models on Human Expression and Thought’. Trends in Cognitive Sciences, 30: 805–816.
Stanley, L. & Wise, S. (1993) Breaking Out Again: Feminist Ontology and Epistemology (2nd edn). London: Routledge.
Turkle, S. (2026) Artificial Intimacy: Who We Become When We Talk to Machines. New York: Little Brown.
Yoo, J. (2025) Writing as Inquiry: Towards Being and Becoming. London: Routledge.

