Critical AI Literacy for Historians

Welcome

This repository provides structured, hands-on exercises for historians to develop critical AI literacy, digital source criticism, and scholarly practice with large language models (LLMs).

WarningDISCLAIMER

This project is currently under active development (with the help of AI) and in a very early stage. The materials provided here are preliminary and may change significantly. Please use them with caution and be aware that they may not yet represent the final quality or scope intended for this resource.

Purpose

These materials aim to help learners:

  • Understand how AI functions as both a research tool and a method
  • Reflect on how it shapes historical interpretation and evidence
  • Apply rigorous digital source criticism to assess data provenance, representation, and bias
  • Practice responsible, transparent, and ethically informed use of AI in historical research

Principles

The exercises promote:

  • Reproducibility in research workflows
  • Data protection and ethical data handling. In its section “What you need”, each exercise names the material that stays outside the AI system, and the rules of data protection law that apply.
  • Copyright awareness and proper attribution
  • Sustainability through minimal computing, that is through procedures with a low demand for computing power. Three properties help here, and each one carries its own condition. A small model for a narrow task needs clearly less energy per request than a large general model (Luccioni et al. 2024). An open-weight model lets you keep one fixed version (Spirling 2023). A local model leaves your material on your computer. The section Check and classify capabilities shows when this way pays off.
  • Interdisciplinary collaboration and critical reflection

By engaging with these materials, historians will be better equipped to design transparent, meaningful projects that integrate AI into their research without compromising disciplinary rigor.

Choose Your Path

Depending on your interests or learning goals, you can follow different paths through the materials. We recommend starting with the Prompt Engineering exercise.

This track guides you through the entire research process using the example of Swiss diplomatic history (Dodis) and the Council of Europe.

  1. Prompt Engineering: Basics of interacting with LLMs.
  2. Research Question: Developing a historical research question.
  3. Source Search: Targeted search for primary sources.
  4. Source Criticism: Applying digital source criticism.
  5. Writing: Drafting and refining historical texts.
  6. Citing: Properly documenting AI use.

This track focuses on literature research and communicating historical topics to the public.

  1. Prompt Engineering: Basics of interacting with LLMs.
  2. Literature Research: Finding and evaluating secondary literature.
  3. Research Question: Formulating a research question. The exercise uses the Council of Europe as its example. Transfer each step to the topic of social security.
  4. Writing: Drafting and refining historical texts.
  5. Citing: Properly documenting AI use.
  6. Public History: Presenting history for a broad audience.

Utilize the methods and templates for your own research project.

  • Prerequisite: Complete the Prompt Engineering exercise first.
  • Transfer: Use the prompts and reflection questions from the other exercises and apply them to your own sources and topics.

All Exercises

Prompt Engineering
Historical work with LLMs: controlled heuristics, reproducibility, evidence discipline
AI & Digital Methods

An exercise that trains central prompting techniques (basics, roles, zero-shot and few-shot, structured reasoning trace, iterative prompting, optimization) along historical…

Literature Research and Analysis
The History of Social Security in Switzerland
Methods

A guided exercise on systematic literature research and problem-oriented literature analysis, with a critically controlled use of generative AI.

Formulating a Historical Research Question
Switzerland and the Council of Europe
Methods

A guided exercise on the systematic development of an analytical historical research question, with a considered use of generative AI.

Source Search for a Historical Research Question
The Vorort and the Accession to the Council of Europe in 1963
Sources

A guided exercise on systematic source search (primary, secondary, tertiary) with a considered use of generative AI.

Source Criticism
Federal Council and Council of Europe 1949
Sources

A guided exercise on systematic external and internal source criticism of a diplomatic memorandum (Dodis).

Academic Writing
Role-based exercise set along the writing process
Methods

An exercise that trains the use of generative AI as a writing assistant in several roles, with a focus on traceability, argument logic and linguistic precision.

Academic Citation
Exercise on citation practice in historical research
Methods

This exercise teaches the fundamentals of academic citation in historical research, with a particular focus on the critical use of generative AI. You learn how citations…

Public History and Historical Communication
Knowledge communication along the writing and production process
Public History

An exercise set that communicates a scholarly article to a target group in several formats (blog, podcast, social media, poster, interview), with an AI log, a source veto…
No matching items

About the author

Moritz Mähr (Dr. sc. ETH Zurich) is an associate researcher in Digital Humanities at the University of Bern and an information and library science specialist with Research Analytics Services at ETH Zurich. His work bridges digital history, science and technology studies, and open research infrastructure, with a focus on digital source criticism in the age of AI, the history of digitization in public administration, minimal-computing approaches to public history, and digital sustainability grounded in FAIR and CARE principles.

About This Project

Critical AI Literacy for Historians is an educational initiative designed to equip historians with the knowledge and skills to critically engage with artificial intelligence technologies in their research practice. The frameworks and approaches presented here are largely based on the working paper Implementing Generative AI in the Historical Studies. (Oberbichler and Petz 2025)

Pedagogical Framework

The project is grounded in the following pedagogical principles:

  1. AI as Tool and Method: Understanding AI not only as a technical tool but also as a methodological approach that shapes research questions and interpretations
  2. Digital Source Criticism: Applying rigorous critical evaluation to AI-generated content, datasets, and algorithmic outputs
  3. Ethical Awareness: Fostering responsible use of AI that considers privacy, bias, representation, and power dynamics
  4. Transparency and Reproducibility: Promoting open, documented, and replicable research practices
  5. Disciplinary Rigor: Maintaining historical methods and critical thinking while integrating new technologies

Each exercise includes

  • Clear learning objectives
  • Prerequisites
  • Hands-on activities
  • Critical reflection prompts
  • Additional resources

FAIR and CARE Principles

This project adheres to:

  • FAIR principles (Findable, Accessible, Interoperable, Reusable) for research data and educational materials
  • CARE principles (Collective Benefit, Authority to Control, Responsibility, Ethics) for responsible data governance

Contributing

We welcome contributions from historians, educators, and researchers. Please see our contributing guidelines for more information.

License

Contact

For questions or feedback, please open an issue or contact the maintainers.

Back to top

References

Luccioni, Sasha, Yacine Jernite, and Emma Strubell. 2024. “Power Hungry Processing: Watts Driving the Cost of AI Deployment?” Proceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparency (FAccT ’24) (New York, NY, USA), June 3, 85–99. https://doi.org/10.1145/3630106.3658542.
Oberbichler, Sarah, and Cindarella Petz. 2025. Working Paper: Implementing Generative AI in the Historical Studies. Version 1.0. February 25. https://doi.org/10.5281/zenodo.14924737.
Spirling, Arthur. 2023. “Why Open-Source Generative AI Models Are an Ethical Way Forward for Science.” Nature 616 (7957): 413. https://doi.org/10.1038/d41586-023-01295-4.