For my research
ia |
Generative AI in research: the European ground rules
Researchers using GenAI in their scientific work can situate that use within the European reference framework and know when to declare their reliance on generative AI.
In May 2026 the European Commission published the third version of its guidelines on the responsible use of generative AI in research. Non-binding, they serve as a common reference for research organisations and European research funders.
Six ground rules apply directly to researchers:
- You remain responsible for the content produced. An AI system is neither an author nor a co-author.
- Declare substantial use in the methods section. Basic editorial support does not count; a literature review, the interpretation of a data analysis or the formulation of hypotheses do.
- Do not upload unpublished or sensitive work to a tool without assurances about how the data will be reused.
- Check the citations: a reference can be accurate and still come with an incorrect summary.
- Respect copyright and the GDPR, on outputs as well as inputs.
- Refrain from using AI to assess a paper or a funding proposal.
Piccolo: UCLouvain's AI applications
Piccolo conversations are encrypted and cannot be used to train commercial models (LLMs).
Piccolo Transcription or Piccolo Projets are the two AI applications most used by researchers at UCLouvain.