This post was automatically translated from the German original at
Digital Lehren und Studieren | Beitrag zur Ringvorlesung des UOS Zentrums für Lehrkräftebildung.
History × Digital History × AI
On 20 January 2026, Christoph Rass spoke as part of the lecture series “Teaching and Learning in the Digital World” at the University of Osnabrück about the potentials and challenges of Artificial Intelligence for historical studies.
The central thesis of the contribution: AI does not change what history is. However, it forces us to articulate more clearly what constitutes historical competence. Building on Michel-Rolph Trouillot’s theory of the four moments in which silence enters the production of history, the lecture developed the concept of a “fifth moment” – that of algorithmic silencing. Training data on which language models are based reproduce existing power relations. The algorithms that select and synthesise information can reinforce existing silencings by reproducing hegemonic narratives and further marginalising marginalised perspectives.
Against this background, the contribution advocated for critical integration rather than blanket rejection or uncritical adoption. The guiding principle: augmentation, not automation – AI should work with historians, not in their place. This requires new competencies: analytical understanding of how language models function, methodological skills for critically examining AI outputs, and reflexive competence for contextualising the social and epistemic consequences of these technologies.
The lecture also presented the Digital History practice of the Modern History and Historical Migration Research working group – from data-driven research projects such as the analysis of the Osnabrück Gestapo card index and the foreigners’ registration card index to the Digital History Workshop, in which students are introduced to digital historical methods in a practice-oriented manner.
On the NGHM website, under “Studies & Teaching” in the section NGHM DH Workshop, further materials can be found: guidance on the use of LLM and AI in academic work, an optimistic and a pessimistic perspective on AI in historical studies, as well as reflections on the future of historical practice under digital conditions.