Professional writing with GenAI in business, universities and the public sector

In many professions, writing is both an outcome and an instrument of professional work. Professionals do not simply write after they know; they come to know, decide, and act through writing. Case notes, patient journals, reports, assessments and emails are not just records of expertise but key sites where professional judgment is formed, negotiated, and made accountable. If writing is central to how professionals learn to become professionals and practice their professional expertise, what happens when writing itself is partly delegated to generative AI (GenAI)? And to what extent might this be context-dependent? This seminar explores ethnographic studies of writing-in-practice and asks what is at stake when generative AI enters these processes: How might professional learning, responsibility, and expertise be reshaped when writing is mediated by GenAI? How are professional practitioners’ sense of their own expertise, professional ethos, agency and meaningfulness transformed when they use GenAI?

With the wide availability of GenAI tools, many workplaces are introducing AI-augmented writing into their work processes. The use of GenAI is addressed in many studies, but less attention has been given to the real-life usage of such tools in professional writing practices and in textually mediated work. There is a need to more extensively explore how the literacy practices of various kinds of professional writers are transforming due to technological developments. We need to understand literacy practices in different kinds of professional workflows and across industries. For instance, how does GenAI-augmented writing differ in the business context compared to the public sector? And how do possible domain-specific nuances map onto questions of ethics, responsibility and agency? Discussion on how GenAI affects the small-scale, everyday and mundane ways of doing things is only beginning to emerge and needs to be brought into dialogue with the current body of research looking into the use and effects of AI in educational contexts. Such a holistic understanding is needed as these tools become more widespread and begin to affect an even wider range of textually mediated practices.

A special issue in Reading Research Quarterly demonstrates the need to begin to develop a literacy studies perspective on studying the use of GenAI (Robinson and Hollet, 2024; Kumar et al., 2024). While the current body of work in literacy studies mostly focuses on student writing, which is justified as higher education institutions are widely affected by the early stages of GenAI implementation, this perspective must urgently be extended to include professional writing. Students are learning professional writing, and their ideas of good texts, genre and register conventions and ethics of text production are still developing. Research on GenAI needs to cover varied forms of writing practices across multiple contexts to gain a better understanding of the scale of the transformations. 

This seminar brings together scholars across literacy studies, education, STS, anthropology and other fields whose research investigates professional and other forms of writing in business, universities and the public sector.

We invite participants to engage with the following questions:
- How do professionals use GenAI? What kinds of writing do they use it for?
- Are there tasks where GenAI works better than others? Or tasks in which GenAI does not work at all?
- Are there differences in the ways students, teachers and other professionals use GenAI? What are the differences between using GenAI for routine tasks and for learning new skills? And what activities are considered ‘routine’?
- How do professionals ‘manage’ GenAI? (E.g. monitoring output, prompt engineering, modifying off-the-self systems?)
- To what extent are professionals developing domain-specific ‘micro-workflows’ to bypass GenAI limitations?
- How has writing changed the way professionals work?

References
Kumar, P. C., Cotter, K., & Cabrera, L. Y. (2024). Taking responsibility for meaning and mattering: An agential realist approach to generative AI and literacy. Reading Research Quarterly 59(4), pp. 570-578.  https://doi.org/10.1002/rrq.570 
Robinson, B., & Hollett, T. (2024). Literacy in the age of AI. Reading Research Quarterly, 59(4), pp. 555–559. https://doi.org/10.1002/rrq.581

Program

13.00-13.05 Opening words by Maja Hojer Bruun

13.05-13.25 Hanna-Mari Pienimäki: The use of generative AI in professional writing: Negotiating textually mediated expertise in hybrid systems 

13.25-13.45 Mads Rosendahl Thomsen: From bottom-up to top-down?

13.45-14.05 Matilde Høybye-Mortensen and Marie Leth Meilvang: Get it wrAIte. The introduction of AI generated minutes in social work

15.40-16.00 Discussion

Abstracts

Hanna-Mari Pienimäki: The use of generative AI in professional writing: Negotiating textually mediated expertise in hybrid systems
In my presentation will present a methodological proposal for studying writing with GenAI and preliminary findings of a four-year (2026–2029) project Disruption in text production? The use of generative AI in professional writing (GAI-WRITE). I analyze business and academic writing case examples to interrogate forms of expertise that resist algorithmic capture and show how professionals, more and less successfully, combine their own non-codifiable, local and tacit expertise with the affordances of GenAI to produce texts. The analysis demonstrates how expertise is distributed and negotiated within human-AI hybrid systems, and how this distribution of labor maps onto the linguistic level.

Hanna-Mari Pienimäki is a postdoctoral researcher and Academy of Finland Research Fellow at the Department of Languages, Helsinki University.

Mads Rosendahl Thomsen: From bottom-up to top-down?
TEXT's 2025 Fall survey of AI use among writing professionals showed a widespread but unsystematic use of AI. Many tasks were frequently carried out with the help of AI, as the use of AI in 2025 became normal for most people, but a majority struggled to get the desired results in their writing. The respondents foresaw an increased pressure to use AI, but this was not matched by training and standards. In this presentation, I present further results from the analysis and ask whether an increased demand for the use of AI tools, though likely, is possibly premature.

Mads Rosendahl Thomsen is head of the research center TEXT: Center for Contemporary Cultures of Text and professor of Comparative Literature, Aarhus University.

Matilde Høybye-Mortensen and Marie Leth Meilvang: Get it wrAIte. The introduction of AI generated minutes in social work
In welfare services all around the world, high hopes are linked to digital technologies as a way to enhance the effectiveness of public administration. The idea is that such technologies may ease – or even take over - administrative tasks, facilitate coordination and uniformity and produce relevant management information. Social work has been no exception and has incorporated technologies such as case management systems, communication systems, automated processes and lately generative AI systems. 

Currently, AI-solutions which record, transcribe, and condensate meetings between professional case workers and clients are rolled out in many Danish municipalities. The potential benefits are more eye contact during meetings, less time spent on documentation after meetings, more well-written minutes, and less biased representation of client-professional encounters. The potential dangers are uncritical import of AI generated products, deskilling of case workers’ writing and reflective skills, distorted representation of client-professional encounter, and invisible management of professional discretion by prompts. 

By following the introduction of AI in a Danish municipality, this project investigates how generative AI influences the writing process amongst social workers. When writing is understood as a professional task that shapes professional thinking and judgments, in what way does AI writing alter professional work and work processes, both written and non-written?

Matilde Høybye-Mortensen is senior associate professor at VIA University College.
Marie Leth Meilvang is associate professor at UCL Erhvervsakademi & Professionshøjskole.

Charlotta Hilli: Historical, affective and sociomaterial trajectories of students’ and GenAI academic writing – perspectives gained from the CO-WRITE project 
The presentation discusses insights gained from the project co-write at Åbo Akademi University, building on two perspectives. First, the presentation addresses workshops with students who have generated data about who and what are involved in their academic writing (e.g., bachelor's thesis, essays, and course assignments). In sociomaterial analyses, we traced human (e.g., supervisors, parents), more-than-human (e.g., digital dictionaries, GenAI) and affective (trust/distrust, ethical concerns) actors shaping students’ academic writing. Second, the presentation addresses an exploration of the historical basis for academic writing in universities. Particularly in the wake of large language models like GenAI, we note a shift towards combining written and oral presentations, thus, in a way, returning to the oral pedagogy of early modern universities in Europe. The project highlights the multifaceted spaces of academic writing, always transforming by connecting to numerous actors.

Project link: https://www.abo.fi/en/project/co-write-collaborative-academic-writing-in-hybrid-learning-spaces-in-higher-education/

Charlotta Hilli is lecturer and associate professor at Åbo Akademi University.

Kasper Engholm Jelby: Socialising with AI: Academic literacy in the age of generative artificial intelligence

The rise of technologies based on generative artificial intelligence is fundamentally changing university students’ academic literacy practices, with implications for the ways in which they are socialised into disciplinary discourse communities. Drawing on data generated through linguistic ethnographic fieldwork among undergraduate health-science students at a Danish university in the spring semester of 2025, my presentation explores how generative artificial intelligence tools are integrated into students’ situated academic literacy practices. By triangulating analysis of data generated through sound, video and screen recordings from participant observation of students’ group work, I argue that students assign to GenAI interlocutors interactional roles previously reserved for (human) peers. Based on this analysis, I suggest that academic literacy socialisation in the age of generative artificial intelligence involves the ability to navigate dialogic spaces inhabited by both human and GenAI actants.

Kasper Engholm Jelby is a PhD Fellow at the University of Copenhagen as part of the sociolinguistic research project ‘AI and the University’ (www.ai-uni.dk), based at the Centre for Internationalisation and Parallel Language Use.

Samantha Goodchild & Jens Christian Borup Green Jensen: Negotiating ideologies of efficiency and thoroughness: STEM and social sciences researchers’ positionings and their GenAI (non)use

Among the burgeoning research on text-generative AI (GenAI) use in universities there has been a predominant focus on teaching and learning. Less attention has been paid to the effects GenAI has on research, based on reported and documented research practice. We contribute to this growing field with two linguistic ethnographic case studies. Investigating the relationship between researchers’ reported practices and their observed practices, we ask if, and how, the use of GenAI affects researchers’ professional identities. Using data gathered from the social sciences and STEM from two universities, we present a critical discourse analysis of interviews focussing on the researchers’ own articulations of their use of GenAI. We then complement these findings to screen recordings of researchers’ GenAI use to show how identity work is being negotiated in the face of the affordances and constraints of GenAI. Using both emic and etic categories such as ‘efficiency’ and ‘thoroughness’, our analyses show the researchers engaging in ideological positioning work (Gal & Irvine 2019), articulating various nested professional identities as experts, researchers, artists, and writers; all of which are up for (re)negotiation when considering the (non)use of GenAI in current academic research. Finally, we situate our findings within wider ongoing discourses on speed, productivity and integrity in the academy, and how these potentially are being affected by the use of GenAI. 

Samantha Goodchild is a postdoctoral researcher and Jens Christian Borup Green Jensen is a PhD fellow, both working on the AI-UNI project (www.ai-uni.dk) at the Centre for Internationalisation and Parallel Language Use, the University of Copenhagen.