Generative AI and the Generous Historian

by | Nov 16, 2026 | AI and History Teaching in UK Higher Education | 0 comments

 

 

This post is part of a series commissioned by the Royal Historical Society on ‘AI and History Teaching in UK Higher Education’ in Autumn 2026.

The series includes a range of perspectives on the uses and limitations of Generative AI and LLMs from history educators and research academics.

Participating authors provide ideas and suggestions for teaching practices, shaped by their own scholarship and teaching experiences in contribution to the ongoing conversation about the impacts of AI on teaching and learning history at UK universities.

 

 

By Dr Adam Budd

When I welcomed my new cohort of first-year History students in October, I said a few words about generative AI: “Using AI apps to write is like using a machine to lift weights. The machine will ease the weights, and it will ease your struggle with the blank page—but it won’t build muscles or knowledge.” What I meant to say is that we study history both to gain knowledge and to learn skills. Using AI to compose text lets us displayknowledge, but it won’t teach us new skills. I agree with Heidi Estrem that writing is a knowledge-making activity and that cutting and pasting is not. AdvanceHE has compiled evidence to show that using AI can weaken students’ learning. Over time, reliance on AI for problem-solving can damage their capacity to learn, exacting “cognitive costs” on their young brains.

I realise now that I said the wrong thing—or the ungenerous thing. A recent report by History UK suggests that students are using AI apps not because they are lazy, but because they recognise the importance of acquiring technical skills for AI-enabled careers. Since historians like me are not helping them to develop those skills, students are experimenting with the technology through their degrees. They do this despite our typical advice, that they must not use AI for their assessments, while we threaten to charge them with academic misconduct if they do so. I went so far as to cite a report that describes AI as a threat to their health.

It is no longer reasonable, let alone helpful, to ask students to set AI aside in their work. Since I cannot do that, can they? As a lecturer at Edinburgh, my employer requires me to use Microsoft 365, so avoiding AI is far harder than it was even a year ago. Without warning or permission, the Copilot app has woven itself into every act of reading and writing that I do on a keyboard. I can’t even set up an office-hours booking page without the software announcing my availability to myself and to others. There’s no opt-out: our data is commercially valuable, and Microsoft isn’t a charity. Even our scholarly databases, which are essential resources for research, “tailor” search results to inferred “preferences.” Logging out simply cuts off access altogether. There are compelling reasons why AI is here to stay and why asking students to put it away while undertaking coursework only reveals obliviousness. (I am not oblivious to the environmental consequences of the technology, which are borne largely by those least able to cope with them. We have an ethical responsibility to face up to this unfairness.)

Our challenge as educators lies in making clear to students that this powerful and pervasive technology may tempt them away from learning the very skills that AI-enabled workplaces will demand of them. But those who anticipate the future job market may have read the World Economic Forum’s projections: AI-driven disruption to the global workforce will entail the loss of some 92 million jobs by 2030 (i.e., over the next three and a half years). Of the 170 million new jobs created in that same period, a solid proportion will demand technical skills but also “human skills, such as creative thinking, resilience, flexibility and agility.” We have seen that integrating AI into our students’ lives can leave them weaker as creative thinkers—less resilient, less flexible, less agile. So it is worrisome that, according to HEPI, 94% of students use GenAI for help with their assessments. Yet, again, half of those students say that their universities do not help them use AI as a tool to support their learning. It is this appetite for effective learning that will guard against cognitive costs: these students are looking for solutions; they are not finding them within our existing pedagogy. We may not be the right people to teach them to use AI effectively—and that may be our primary problem. This should come as no surprise. How many of us completed our first degrees before JSTOR was launched in 1995? How many of us still struggle to share our screens, five years after the lockdowns made that a critical teaching technique? Still, it is we who define academic misconduct, we who design History degrees, and we who assess them. This may need to change. We must offer a curriculum that meets our students’ needs. This will mean taking the brave step of collaborating beyond our field on core skills training.

This is the critical task historians face. Employers will require skills with AI that include critical thinking, reflective discussion, thoughtful and persuasive analysis based on high-quality evidence. These are the very skills that have defined historical education at universities since Ranke brought his students out of the lecture-hall and assembled them around a table, some 200 years ago. Back then, most students of history were affluent young men seeking Bildung, or moral self-realisation. Relatively few sought skills for the workplace. Our students need skills for roles that require both technical and historical abilities, in a combination we have not yet seen. The way AI works remains mysterious, even for its greatest engineers. But it is our task to understand what AI asks of us as educators. We must not allow its apparent easing of effort to seduce our students away from building new skills. We must respond to this socio-technical change generously, which means collaboratively, retaining our values as scholars who prize integrity, diligence, and critical analysis and integrating expertise we might not have. We must ensure that AI can help prepare our students for the future, so that the future will not define them.

 

Recommendations for Further Reading

M. Aczel, et al. Environmental Cost of AI’s Energy Use: Carbon, Water and Land Footprints. United Nations University Institute for Water, Environment and Health (UNU-INWEH), 2026. doi: 10.53328/INR26RMA002

D. Amodei, “The Urgency of Interpretability” (April 2025).

Heidi Estrem, “Writing is a Knowledge-Making Activity,” in Linda Adler-Kassner and Elizabeth Wardle (eds.), Naming What We Know: Threshold Concepts of Writing Studies (Logan, UT: Utah State University Press, 2015), 19-20. https://scholarworks.boisestate.edu/english_facpubs/268/

History UK, “History and AI” project. https://www.history-uk.ac.uk/projects/history-and-ai/

Elizabeth Kaplunov and Athina Ntasioti, “Is AI making students less critical thinkers?”, Advance HE, 22 September 2025. https://advance-he.ac.uk/news-and-views/ai-making-students-less-critical-thinkers/

R. Stephenson and C. Armstrong, Student Generative Artificial Intelligence Survey 2026 (Oxford: Higher Education Policy Institute, 2026).

T. D. Wilson, “The Development of Policies on Generative Artificial Intelligence in UK Universities,” IFLA Journal 51 (2025). https://journals.sagepub.com/doi/10.1177/03400352251333796

World Economic Forum, Future of Jobs Report 2025 (Geneva: World Economic Forum, January 2025).

 

About the Author

Dr Adam Budd is Secretary for Education at the Royal Historical Society, and Senior Lecturer in Cultural History at the University of Edinburgh.

Follow This Blog

Subscribe

* indicates required



Categories

Follow us on BlueSky