Many apply; most are are denied - but for Baraa Khuder and Raffaella Negretti at the Department of Communication and Learning in Science it was a jackpot. The Swedish Research Council has awarded nearly five million SEK for their project Knowledge making in the era of Artificial Intelligence: social regulation of writing for publication processes and products.
Would you trust a scientific article that was partly written with the help of artificial intelligence? As AI becomes increasingly used in research writing — especially among junior researchers — new and important questions are being raised about integrity, diversity, and reliability in academic texts. These are exactly the issues at the heart of the project “Knowledge making in the era of Artificial Intelligence,” which has recently received funding from the Swedish Research Council, as part of a strategic call for projects on the social consequences of digitalization.
“The decision came at 2:00 p.m., and at 2:01 I was on the phone with Raffaella screaming out loud, ‘We got it!’ It feels unreal and incredibly exciting, because we have put so much thought and care into this project,” says Baraa Khuder.
“I’ve been through the Swedish Research Council’s application process before and honestly didn’t have any expectations at all. It was a huge shock, and there were happy dances, hugs, and tears in the office for quite a while. The success rate is 7 percent! I still can’t believe we got it! At the same time, it feels so good to see that all the hard work has paid off,” says Raffaella Negretti.
“Our success is also a recognition of the collegiality we have within the department and of the value of collegial support and collaboration. It also shows the internal process for grant applications that we have established at the department works. Given Chalmers’ excellence initiative, this is very positive not only for us but for the entire university”.
The background to the project is the researchers’ concern that today’s reliance on AI-generated language — often trained on dominant norms — may unintentionally crowd out alternative ways of expressing knowledge, and that interaction with AI may diminish the author’s agency over their scientific articles.
“How we write shapes how knowledge is understood, used, and trusted, both within academia and in society at large,” says Baraa Khuder.
Junior researchers also face strong publication pressures and may therefore turn to AI tools for support with argumentation, text structure, and language.
“These tools can be useful, but we still know surprisingly little about how they affect the writing process itself and the finished text. Does AI-assisted writing lead to more standardized, less critically reflective, and less stylistically varied research articles? And if so, what does that mean for the breadth, creativity, and credibility of research fields?” says Raffaella Negretti.
In this interdisciplinary project, composed of Raffaella Negretti, PI, Baraa Khuder, co-applicant, Moa Johansson, co-applicant from the Department for Computer Science and Engineering, and Theresa Lillis, external adviser, nine junior researchers from different disciplines will be followed over time. Using digital tools and recurring interviews, the team will investigate how they use AI when writing scientific texts, what choices they make, and what consequences that use has. The research group will also analyze published articles to see whether AI use can be linked to a more uniform and less varied writing style.
The goal is to contribute new knowledge about how digitalization affects the content and forms of expression in research. The results are expected to provide input for policy development, educational initiatives, and the design of future AI tools — so that digitalization supports, rather than undermines, integrity, diversity, and reliability in scientific texts.
How will you celebrate?
“We’re going to celebrate by dancing! But Raffaella doesn’t know that yet,” laughs Baraa Khuder.
On our department website you can find more about our research related to AI and higher education