Conference Agenda

This symposium seeks to critically examine uses of AI in educational research beyond hyped and hegemonic discourses that peddle AI as ‘inevitable’, a mere ‘tool’, or an ‘opportunity’ that cannot be missed within academic practices (Nemorin et al., 2022). We explore how AI is affecting different spheres of educational research, including research training, research design and development, as well as dissemination of findings. In broad terms, we address the AI hype by attempting to escape the usual good-bad dichotomy and envisage alternative ways of thinking about artificial intelligence in education (AIED). As critical scholarship has pointed out, this entails exploring political, affective, cognitive, and ethical implications of AI, going beyond mainstream concerns with academic integrity and plagiarism (Selwyn 2024; Williamson et al., 2023).

The use of AI has proliferated rapidly in academia since the advent of powerful Large Language Models (LLMs), generating enthusiasm and fear alike (Driessens & Pischetola, 2024). Of particular concern are issues that pertain to educational research, including originality and authorship, data security, confidentiality, but also impacts on cognition, critical thinking, creativity, as well as reading and writing skills (Atkinson & Flanagan, 2024). LLMs are also becoming de facto search engines for academics, with impacts on the relevance and accuracy in information outputs (Haque & Li, 2025). How is educational research affected by the popularisation of chatbots as ‘epistemic authority’? What are the affects connected to AI uses in everyday scholarly activities? What are professional and ethical implications for academic work?

The symposium will bring together contributions from researchers engaged in reflection around these issues as they take shape within different countries: UK, Brazil, Denmark, Sweden, and Australia.

Paper one, by Francesca Peruzzo and Marina Avelar, employs affect as an analytics to understand how AI policies are enacted, avoided, or left ambiguous within educational research. Drawing on their situated scholarly experiences in the United Kingdom and Brazil, which take the shape of six vignettes in the paper, the authors reframe AI policy enactment as an affective and situated process irreducibly shaped by power, place and positionality.

In paper two, Edgar Lyra explores the issue of genAI’s ‘epistemic authority’ and the prior conditions of possibility for this status quo, e.g. scientific publishing, academic productivism and questionable referencing practices for text validation. The paper discusses the naturalisation of these practices, both in Brazil and globally, and the historical processes that have allowed for a general acceptance of chatbots in academic research.

Paper three, by Magda Pischetola and Giselle Ferreira, delves into the increasing use of automation in the context of research training, supported by arguments of efficiency and time, and framed as desirable for academic freedom. By drawing on concrete examples from Denmark and Brazil, the authors argue that research training automation is neither efficient nor desirable, as it not only hinders development of key research skills but also feeds precisely the neoliberal paradigm it assumedly circumvents.

Finally, paper four, by Anders Sonesson and colleagues, focuses on the intrinsic ethos of academic work, and explores what choices around genAI uses make scholars feel uneasy or reticent about their practices. Drawing on preliminary findings from a survey that gathered ‘confessions’ of scholars from different countries (Sweden, Australia, Italy, Canada, among others), the contribution addresses ‘grey areas’ of genAI use perceived as being at the limit of ethical boundaries.

These four perspectives allow for a nuanced and critical discussion around the uses of AI in educational research, providing not only a range of different theoretical frames but also empirical data and case studies that are used by the authors to question, dismantle, and overcome the hype around AI.

This paper examines how academic researchers engage with Generative Artificial Intelligence (GenAI) not from neutral positions but from within specific affective, institutional and geopolitical conditions. Drawing on autoethnographic dialogue between two scholars based in the United Kingdom and Brazil, we trace an affective trajectory through which encounters with AI move from techno-solutionist excitement, through disillusionment, to practices of refusal. We argue that this trajectory is neither incidental nor merely personal but is patterned by the conditions of academic labour and the colonial matrix of power within which GenAI circulates. To analyse these dynamics, we weave together three interlocking theoretical frameworks. We mobilise affect (Massumi, 2002; Ahmed, 2004) as an analytic to examine how relational intensities are diagnostic of the systems within which AI is encountered. We situate these affective experiences within decolonial perspectives on knowledge production to show how GenAI reproduces existing epistemic hierarchies, defaulting to Global North data, dominant languages and Eurocentric frameworks in ways that are felt differently across Global North and South contexts. We further draw on neo-material assemblage thinking (Stark, 2019; Grimaldi and Ball, 2021; Perrotta et al., 2024; Landri and Peruzzo, forthcoming) and the concept of cruel optimism to theorise how AI operates as an object of promise that simultaneously reproduces the performative conditions it appears to relieve. Our analysis is developed through six vignettes grounded in our everyday research practices, which offer situated insights into how affective engagements with AI emerge within specific institutional and geopolitical contexts. In conclusion, we propose five conceptual tools for researchers working at the intersection of AI and higher education, including affect as a diagnostic, assemblage thinking as method, cruel optimism as critical approach, coloniality as necessary context, and refusal as a legitimate scholarly practice. Together, these tools reframe AI policy enactment as an affective and situated process irreducibly shaped by power, place and positionality (Williamson et al., 2023; Selwyn, 2024).

Ahmed, S. (2004). The Cultural Politics of Emotion. New York: Routledge. Grimaldi E. and Ball, S. J. (2021). The blended learner: digitalisation and regulated freedom - neoliberalism in the classroom, Journal of Education Policy, 36:3, 393-416. Landri, P. and Peruzzo, F. (forthcoming-2026). AI and Education: a more-than human approach to education governance. In Lesley Gourlay et al (Eds ). The Palgrave Handbook of Science and Technology Studies in Education. Palgrave MacMillan. Massoumi, B. 2002. Parables for the virtual: movement, affect, sensation. Durham and London: Duke University Press. Perrotta, C.; Selwyn, N. and Ewin, C. (2024). Artificial intelligence and the affective labour of understanding: The intimate moderation of a language model. New Media & Society 26(3), 1585-1609. Selwyn, N. (2024). On the Limits of Artificial Intelligence (AI) in Education. Nordisk tidsskrift for pedagogikk og kritikk: Special Issue on Artificial Intelligence in Education, 10, 3–14. Stark, L. (2019). In J. Vertesi and D. Ribes, (Eds.) digitalSTS: A Field Guide for Science & Technology Studies. Princeton University Press. Williamson, B., Macgilchrist, F. & Potter, J. (2023). Editorial: Re-examining AI, automation and datafication in education. Learning, Media and Technology, 48(1), 1-5.

Academia is currently grappling with the misuse of generative artificial intelligence in the production of texts and general scientific research. There is a growing number of complaints about methodological adulteration, as well as retractions of recently published texts in specialized journals. There is, at the same time, an abusive use and sharp dissatisfaction with the abundance of publications produced mainly by AI. A worldwide wave of growing distrust in science, even in its most insidious form – sheer denialism – adds to the harshness of the scene. Thus, despite efforts to write guidelines on the responsible use of AI in research, it is urgent to discuss the issue of ‘epistemic authority’, along with the consequences of such distrust for public opinion, and the sustainability of dialogical practices. The hypothesis explored in this presentation is that the conditions of possibility for this state of affairs date back to times before the hype of generative AI, that is, to a publishing economy based on productivism, on questionable practices of reference and textual validation; in other words, on a merely indexical textuality, in which the real examination of sources and their actual content are humanly impossible, spuriously simulating authoritative validation. In short, we aim to reflect on the naturalization of these practices and on their contribution to the present general lenience and admissibility of generative black boxes.

European Commission (2025). Living guidelines on the responsible use of generative AI in research. European Union, second edition. Messeri, L. and Crockett, M. (2024). Artificial intelligence and illusions of understanding in scientific research. Nature, 627, 49-58. Lyra, E. and Souza, C. (2025). Less Visible Risks of AI-based Technologies: the civilizational threat of discursive and epistemic dysfunction. In Tarantola, A. M. (ed): Artificial intelligence and the Care for our Common Home – a focus on Industries, Finance, Education and Communication. Milano, Vita e Pensiero. Popowicz, D. M. (2024). The epistemic authority of practice. In M. Farina, A. Lavazza (Eds.). Philosophy, Expertise, and the Myth of Neutrality. New York, Routledge. The Retraction Watch Database (Blog) (2026). New York, The Center for Scientific Integrity. Available from http://retractiondatabase.org/ Van Noorden, R. (2023). More than 10,000 research papers were retracted in 2023 – a new record. Nature.

Surveying literature is a key research process that invariably poses challenges to experienced and novice researchers alike. With the exponential increase in numbers of academic publications fuelled by genAI, automation is being peddled as the way forward for dealing with large collections of academic texts. In this context, prompt engineering is rapidly gaining the status of a Holy Grail, especially for aspiring researchers. This paper aims to discuss the use of automation in the context of research training, with focus on the processes of locating, selecting and reviewing literature. The main arguments in support of automating these processes are couched in terms of efficiency and time. These ideas are associated with the expectation that genAI can cover so-called tedious research tasks, suggesting a conception of research as an optimised process driven by the logic of speed, which generates knowledge in the form of quantifiable outputs. This perspective disregards both researchers’ experience and context, and, even more problematically, considers time as a resource to be ‘saved’ and used for more appealing activities. On these grounds, genAI is framed as a desirable support for academic freedom, perhaps as the ultimate opportunity to escape the neoliberal constraints of the contemporary university. As trainee researchers turn to AI platforms to produce ready literature reviews for projects and assignments, we ask, paraphrasing Lewis Mumford (1964): might the ‘colossal benefits’ advertised by AI evangelizers not bring about any type of ‘colossal loss’? By drawing on concrete examples from our experiences in Denmark and Brazil, we analyse how efficient and time-saving prompt engineering requires skills that are yet to be acquired by early-career scholars, and might become the very obstacle to this acquisition. In line with critical studies underlying the potential of slow scholarship, digital degrowth, and the praxis of refusal, we argue that this form of automation in research training is neither efficient nor desirable, as it not only hinders development of key research skills, but, crucially, it feeds precisely the neoliberal paradigm it assumedly circumvents.

Jowsey T. et al. (2025). We Reject the Use of Generative Artificial Intelligence for Reflexive Qualitative Research. Qualitative Inquiry, 2025, https://doi.org/10.1177/10778004251401851 McQuillan, D. (2022). Resisting AI: An anti-fascist approach to artificial intelligence. Policy Press. Mumford, L. (1964). The automation of knowledge. AV Communication Review, 12(3), 261-276. OCDE (2026). OCDE Digital Education Outlook 2026. Exploring Effectives Uses of Generative AI in Education. Paris, OCDE Publishing. Saetra, H. S. (2025). The rise of the research automaton: science as process or product in the era of generative AI? AI and Society, 2025, https://doi.org/10.1007/s00146-025-02557-7 Selwyn, N. (2025). Digital degrowth: Radically rethinking our digital futures. Cambridge, Polity. Wahab, S., Mehrotra, G. R., & Myers, K. E. (2022). Slow scholarship for social work: A praxis of resistance and creativity. Qualitative Social Work, 21(1), 147-159. Watermeyer, R., Phipps, L., Lanclos, D., & Knight, C. (2024). Generative AI and the automating of academia. Postdigital Science and Education, 6(2), 446-466.

This paper explores the ethical dilemmas being raised by the fast-increasing use of generative AI by academics looking to support and/or augment their research, teaching, and administrative work. In contrast to claims of genAI being an unproblematic means of academics now ‘working smarter, not harder’, this paper takes a critical stance on how genAI is altering the ethical basis of academic work. In particular, we take a lead from critical AI studies that have addressed the fundamental difference between human scientific process and AI-generated ‘science-like’ output. This work posits that genAI can, at best, only create an impoverished illusion of academic work – abstractions created from the misappropriated work of other academics. In contrast, human scientific work remains deeply meaningful for individuals and societies. This meaningfulness is rooted in embodied and relational entanglements that are not algorithmically replicable through statistical prediction or encoding of semantic regularities. In this sense, the case of academics choosing to use genAI to support/supplant their own work could be argued to fundamentally challenge the intrinsic ethical dimensions of academic work, and the moral purpose driving it – that is, supporting human/social flourishing. The ethos of academic work goes beyond scientific integrity and plagiarism. It entails care for those who contributed to the intellectual traditions we inherit, responsibility towards our readership, as well as curation of scholarly work that will be shaping future generations’ knowledge. Given all of this, a key question emerges: how can academics justify now using genAI as a key part of their scientific work? Against this backdrop, we explore what choices around genAI uses make scholars feel uneasy or reticent about their practices. Drawing on preliminary findings from an exploratory survey (n=135) that gathered anonymous ‘confessions’ from scholars based in different countries (Australia, Sweden, Italy, Canada, among others), we inquire into emotional aspects reported by academics about everyday decisions around the (non-)use of AI. These include frictions between external and internal expectations and competing feelings that challenge the perceived borderline between the acceptable and the unacceptable. Using virtue ethics as a theoretical framework, and in particular the concept of phronesis, we discuss this ethics as lived and enacted. We approach ethics not as grounded in individual choice but as a collective concern for sustaining the conditions of a diffractive, critically conscious and careful engagement with knowledge production – conditions that include practices and spaces that allow for wonder and wisdom to emerge.

Barad, K. (2007). Diffractions: Differences, Contingencies, and Entanglements That Matter. In Meeting the Universe Halfway: Quantum Physics and the Entanglement of Matter and Meaning (pp. 71–94). Duke University Press. Bender, E. M. (2024). Resisting Dehumanization in the Age of “AI”. Current Directions in Psychological Science, 33(2), 114-120. Haraway, D. J. (1998). Diffraction as Critical Consciousness. In Goodeve, T. N. (Ed.). How like a leaf: an interview with Thyrza Nichols Goodeve, p. 101-108, Routledge. Sætra, H. S. (2025). The rise of the research automaton: science as process or product in the era of generative AI?. AI & Society, 1-15, https://doi.org/10.1007/s00146-025-02557-7 Vallor, S. (2016). Technology and the Virtues. New York: Oxford University Press. Vallor, S. & Ganesh, B. (2023). Artificial Intelligence and the Imperative of Responsibility: Reconceiving AI Governance as Social Care. In Maximilian Kiener (ed.), The Routledge Handbook of Philosophy of Responsibility, p. 1–12, Routledge.