# Critical AI Literacy Society
_This text contains no LLM output_
>“The systems being sold as ‘AI’ are not fit for purpose for educational applications and we should not treat it as a foregone conclusion that they represent the future of education.”
>(Bender, 2025, p. 41)
## Our Mission Statement
- To promote _genuinely_ critical "artificial intelligence" literacy in the University of Roehampton's community.
- By organising workshops and conferences, and publishing materials, showcasing best practices in use of large language models (LLMs). The training offered is following a principle of harms-reduction (cognitive, environmental, social...), and grounded in the latest research in this area.
- By disseminating critical AI scholarship carried out within Roehampton and beyond, raising awareness of hazards, harms, and best practices.
- By countering discourses of technological determinism, technological solutionism, green-washing, and _literacy laundering_, where '“AI literacy” as currently deployed is itself a symptom of the confusion it purports to address' ([Ray and Guest, 2026](https://zenodo.org/records/21222978), p. 1)
- By representing, within the RSU, the concern for the lack of critical reflection in the University's AI strategy and have the RSU adopt a motion similar to that recently [voted by the University of Warwick's Student Union](https://theboar.org/2026/11/breaking-students-vote-in-favour-of-ethical-ai-and-disaffiliating-from-nus/) : “to lobby the university for an ethical stance on AI”.
## Our Core Team
- **President:** Will Berard, *PhD student, School of Education*
- **Secretary:** Abbie Cushion, *PhD student, School of Arts, Humanities and Social Sciences*
- **Treasurer**: Julie Price, PhD student, *School of Psychology*
With the support of fellow post-graduate research students:
- Benoît André, *PhD student, School of Arts, Humanities and Social Sciences*
- Bridget Steenkamp, *DTh student, School of Arts, Humanities and Social Sciences*
- Irene Benewaa, *PhD student, School of Education*
- Imogen Reeve, *PhD student, School of Dance*
- Andy Teye, *PhD student, School of Arts, Humanities and Social Siences*
- YOU🫵
## Manifesto
>“AI not only ends the past, and makes the present frantic, it also forecloses any future. Under AI’s sway, the future is neither bright as the technopositivists promise, nor a descent into rule by robotic overloads, but a constant rehashing of the past, wherein human creativity and communication are not only mediated by controlled by companies. In the midst of this nonsense, we must nourish hope in shared values, skills, and humanity”
>(Guest *et al.*, 2025, p. 8)
The University of Roehampton's Critical Artificial Intelligence Literacy Society (UR-CAILS) was founded by doctoral students, carrying out postgraduate research in varied disciplines: Arts, Education, Humanities, Psychology and Theology. It is a response to a trend which concerns these students gravely: the seemingly acritical adoption of so-called 'generative AI' technologies, within each of their disciplines, and in society at large. A particularly concerning aspect of such acritical adoption is that taking place across all phases of Education in the UK, and perhaps none more so than in Higher Education, our own institution included.
As regards the _applications_ of AI, we are particularly concerned with:
- the harmful effects of Large Language Models (LLMs) on learning, be it that of formal education settings ([Stromberg _et al._, 2026](https://cepr.org/voxeu/columns/generative-ai-learning-penalty-secondary-school)), or lifelong learning in professional life ([Shen and Tamkin, 2026]()).
- the increasing prevalence of LLM output in published research, sometimes un-declared ([Cabanac, Labbé and Magazinov, 2021](https://arxiv.org/abs/2107.06751)), and the use of LLM tools for peer-review ([Sun _et al._, 2026](https://www.nature.com/articles/s42256-026-01349-y)) , or [evaluation under the Research Excellence Framework](https://acss.org.uk/news/watch-has-generative-ai-already-transformed-research-assessment-and-will-it-unlock-the-refs-regulatory-burden/).[^1]
- the harms of AI assistance onto persistence ([Liu _et al._, 2026](https://ora.ox.ac.uk/objects/uuid:52f0b738-d802-49f4-9976-21ce4deb6b66)) and critical thinking skills ([Gerlich, 2025](https://www.mdpi.com/2075-4698/15/1/6); [Lee _et al._, 2025](https://dl.acm.org/doi/full/10.1145/3706598.3713778)).
- the homogenisation of language ([Sourati, Ziabari and Dehghani, 2026]((Sourati, Ziabari and Dehghani, 2026))), thinking ([de Rooij and Biskjaer, 2026](https://osf.io/preprints/psyarxiv/rz5s4_v1)), and creativity ([Wenger and Kenett, 2026](https://doi.org/10.1093/pnasnexus/pgag042)) concomitant to collective LLM use.
- Spirals
- PISA/ED
As regards the _implications_ of AI, we are particularly concerned with:
- model training being based on the pillage of our epistemic commons. The past intellectual labour of our species has been abstracted into the parameters of language and image models ([Bender and Hanna, 2025](https://thecon.ai); [Hao, 2026](https://www.penguin.co.uk/books/460331/empire-of-ai-by-hao-karen/9781802064650)). The AI industry, appropriating out collective efforts for their private profit, is now moving towards the enclosure of these epistemic commons ([Williams, 2026](https://arxiv.org/abs/2610.27251)).
- LLM post-training and fine-tuning using opaque, exploitative labour practices, often in the Global South. Such practices are the knowledge equivalent of the extractive colonialism performed by empires, past and present ([Crawford, 2021](https://yalebooks.co.uk/book/9780300264630/atlas-of-ai/); [Bender and Hanna, 2025](https://thecon.ai); [Hao, 2026](https://www.penguin.co.uk/books/460331/empire-of-ai-by-hao-karen/9781802064650)).
- the multifarious environmental harms (energy, water, noise... ) of the data centres used to train and deploy models ([Crawford, 2021](https://yalebooks.co.uk/book/9780300264630/atlas-of-ai/); [Bender and Hanna, 2025](https://thecon.ai)), and the rush to construct even more such _'Electricity Eating Monsters_ ([@NovaraMedia, 2026](https://www.youtube.com/shorts/unXibfj3L9I)).
- the economic, and societal impact of the AI economic bubble and its inevitable, yet long delayed, correction.
Our approach is that of Critical AI Literacy in the conception of Guest (2026), where non-use is the default position, and use should prioritise local model, or cloud compute giving visibility over energy use. This use should be critical, reflexive, deliberate, defensible (*e.g.* [Vendrell and Johnston, 2026](https://www.sciencedirect.com/science/article/pii/S2666920X26000342)) , and disclosed to those concerned: typically, your lecturer, examiner, student, or colleague. Use of AI should never be mandated, and non-users have the same right to an inclusive education as AI users. AI users, conversely, deserve explicit, research-informed, critical training, which promotes more ethical alternatives to the systems of Big Chatbot. We want to serve both groups, and get them in dialogue.
---
[^1]: The low barrier to entry to the production of text, often more that can be (and is) meaningfully reviewed, is driving parties at the receiving end to reach for LLMs to assist, augment or outsource reading. LLMs will rate LLM output more highly than authentic human text [(Laurito _et al._, 2025](https://www.pnas.org/doi/abs/10.1073/pnas.2415697122)), which in turns creates a perverse incentive for academic authors to use LLMs. This phenomenon is [already well under way](https://www.rhsmith.umd.edu/news/ai-hiring-tools-may-favor-their-own-work-smith-study-finds) when it comes to CVs and their triage in professional recruitment ([Xu _et al._, 2025](https://arxiv.org/abs/2509.00462)).
## References
@NovaraMedia (2026) _AI Data Centres Are ‘Electricity Eating Monsters’ Supercharging Climate Collapse_. Available at: [https://www.youtube.com/shorts/unXibfj3L9I](https://www.youtube.com/shorts/unXibfj3L9I) (Accessed: 20 November 2026).
Bender, E.M. (2025) ‘We do not have to accept AI (much less GenAI) as inevitable in education’, in UNESCO, _AI and the future of education_, pp. 41–45.
Bender, E.M. and Hanna, A. (2025) _The AI Con: How to Fight Big Tech’s Hype and Create the Future We Want_. Random House.
Cabanac, G., Labbé, C. and Magazinov, A. (2021) ‘Tortured phrases: A dubious writing style emerging in science. Evidence of critical issues affecting established journals’. arXiv. Available at: [https://doi.org/10.48550/arXiv.2107.06751](https://doi.org/10.48550/arXiv.2107.06751).
Crawford, K. (2021) _Atlas of AI: power, politics, and the planetary costs of artificial intelligence_ (1 online resource (327 pages) : illustrations vol). New Haven: Yale University Press. Available at: [https://public.ebookcentral.proquest.com/choice/publicfullrecord.aspx?p=6478659](https://public.ebookcentral.proquest.com/choice/publicfullrecord.aspx?p=6478659) (Accessed: 20 November 2026).
Gerlich, M. (2025) ‘AI Tools in Society: Impacts on Cognitive Offloading and the Future of Critical Thinking’, _Societies_, 15(1), p. 6. Available at: [https://doi.org/10.3390/soc15010006](https://doi.org/10.3390/soc15010006).
Guest, O., Suarez, M. and van Rooij, I. (2025) ‘Towards Critical Artificial Intelligence Literacies’. Available at: [https://doi.org/10.5281/zenodo.17786243](https://doi.org/10.5281/zenodo.17786243).
Hao, K. (2026) _Empire of AI_. Available at: [https://www.penguin.co.uk/books/460331/empire-of-ai-by-hao-karen/9781802064650](https://www.penguin.co.uk/books/460331/empire-of-ai-by-hao-karen/9781802064650) (Accessed: 20 November 2026).
Laurito, W. _et al._ (2025) ‘AI–AI bias: Large language models favor communications generated by large language models’, _Proceedings of the National Academy of Sciences_, 122(31), p. e2415697122. Available at: [https://doi.org/10.1073/pnas.2415697122](https://doi.org/10.1073/pnas.2415697122).
Lee, H.-P.H. _et al._ (2025) ‘The impact of generative AI on critical thinking: Self-reported reductions in cognitive effort and confidence effects from a survey of knowledge workers’.
Liu, G. _et al._ (2026) ‘AI Assistance Reduces Persistence and Hurts Independent Performance’. arXiv. Available at: [https://doi.org/10.48550/arXiv.2604.04721](https://doi.org/10.48550/arXiv.2604.04721).
de Rooij, A. and Biskjaer, M.M. (2026) ‘Does Generative AI Make Us Think Alike? A Systematic Review and Meta-Analysis of Homogenization Effects in Human–AI Co-Creation’. PsyArXiv. Available at: [https://osf.io/preprints/psyarxiv/rz5s4_v1/](https://osf.io/preprints/psyarxiv/rz5s4_v1/) (Accessed: 19 May 2026).
Shen, J.H. and Tamkin, A. (2026) ‘How AI Impacts Skill Formation’. arXiv. Available at: [https://doi.org/10.48550/arXiv.2601.20245](https://doi.org/10.48550/arXiv.2601.20245).
Sourati, Z., Ziabari, A.S. and Dehghani, M. (2026) ‘The homogenizing effect of large language models on human expression and thought’, _Trends in Cognitive Sciences_, 0(0). Available at: [https://doi.org/10.1016/j.tics.2026.01.003](https://doi.org/10.1016/j.tics.2026.01.003).
Stromberg, D., Lei, V. and Wu, Y. (2026) ‘The Generative AI Learning Penalty: Evidence from Chinese Secondary Education’. Rochester, NY: Social Science Research Network. Available at: [https://doi.org/10.2139/ssrn.6868618](https://doi.org/10.2139/ssrn.6868618).
Sun, Y. _et al._ (2026) ‘When AI-assisted criticism scales, peer review needs an obligation threshold’, _Nature Machine Intelligence_, pp. 1–1. Available at: [https://doi.org/10.1038/s42256-026-01349-y](https://doi.org/10.1038/s42256-026-01349-y).
Vendrell, M. and Johnston, S.-K. (2026) ‘Scaffolding critical thinking with generative AI: Design principles for integrating large language models in higher education’, _Computers and Education: Artificial Intelligence_, 10, p. 100572. Available at: [https://doi.org/10.1016/j.caeai.2026.100572](https://doi.org/10.1016/j.caeai.2026.100572).
Wenger, E. and Kenett, Y.N. (2026) ‘Large language models are homogeneously creative’, _PNAS Nexus_, 5(3), p. pgag042. Available at: [https://doi.org/10.1093/pnasnexus/pgag042](https://doi.org/10.1093/pnasnexus/pgag042).
Williams, P.K.G. (2026) ‘AI and The Enclosure of the Scientific Commons’. arXiv. Available at: [https://doi.org/10.48550/arXiv.2610.27251](https://doi.org/10.48550/arXiv.2610.27251).
Xu, J., Li, G. and Jiang, J.Y. (2025) ‘AI Self-preferencing in Algorithmic Hiring: Empirical Evidence and Insights’. arXiv. Available at: [https://doi.org/10.48550/arXiv.2509.00462](https://doi.org/10.48550/arXiv.2509.00462).
# Notes