Breadcrumb

Guide to the Use of Generative AI in Academic Work

Guide to the use of generative AI in academic work

The use of generative Artificial Intelligence (AI) tools in universities is becoming increasingly common. In order to ensure transparency, academic integrity, and proper attribution of authorship, the University of the Basque Country (EHU) establishes that any use of generative AI in undergraduate dissertations (TFG), master’s theses (TFM), doctoral theses, and any assessed work must be explicitly declared and, where appropriate, formally cited in accordance with academic standards.

1. Declaring ≠ Citing: two different actions

Declaring and properly citing the involvement of AI helps to protect student authorship, ensure genuine learning processes, and uphold the university’s academic quality standards.

Declaring the use of AI

Declaring = informing about its use → mandatory if AI has contributed.

✅ AI use must always be declared when it:

  • Generates content that is incorporated into the work, even if rewritten
  • Translates, corrects, or reformulates text included in the document
  • Contributes to the structure, ideas, outlines, or methodology
  • Produces summaries or explanations that are used in the final text
  • Influences the submitted work in a verifiable way

Citing AI

Citing involves formally referencing the tool when AI-generated text (either verbatim or rewritten) is included in the document.

Citing = formal reference → mandatory if AI-generated text appears in the work.

✅ AI must be cited when it:

  • Provides verbatim text included in the work
  • Generates formulations that appear adapted in the final version
  • Produces translations included in the document
  • Creates summaries or explanations reproduced in the text

❌ It is NOT necessary to declare or cite AI when:

  • It is used only to learn or clarify doubts
  • No generated text is incorporated into the document
  • It does not influence the ideas, structure, or writing
  • Its use is purely instrumental: it helps understanding, but it does not appear in the submitted work