In this dossier, we keep track of all relevant material we produce on the topic of generative AI / LLM and their relation to biomedical science, specifically at the LUMC. Given that these materials are produced for different audiences, there can be substantial overlap in the content of each item. They are listed with the newest contributions at the top of the list.
Responsible use of LLMs in research – LUMC course (2026)
In 2026, we developed the course Responsible Use of Large Language Models in Research for researchers at the LUMC. The course moves beyond general introductions to generative AI and focuses on the critical, effective, and responsible use of LLMs throughout the research cycle.
The programme consists of an introductory session that provides the conceptual foundation, followed by practical theme sessions on topics including literature searching, study design, data collection, data analysis, scientific writing, peer review, and grant writing. Particular attention is paid to scientific responsibility, data protection, transparency, validation of LLM-assisted work, and situations in which the use of LLMs is inappropriate.
The course is supported by the LUMC programme Digitaal Dichtbij and forms part of our broader work on AI literacy and the responsible integration of generative AI into biomedical research.
The Leiden University has two websites with relevant information, albeit more related to teaching activities. On this page you can find an overview inclusing a FAQ. This page provides insights from the university library.
The third update on LLM on the LUMC intranet. This update provides an overview of the latest developments and future outlook. Only for LUMC employees.
Overview of actions @ LUMC in the first 6 months of widespread availability of LLM. A presentation held for all policy advisors scientific integrity of Dutch Universities during a meeting in Rotterdam June 2023. A shorter version of this presentation was also presented at UNL (Universities of the Netherlands).
ChatGPT en wetenschap – hoe zit dat eigenlijk? Een beschouwing voor het voorbereidend wetenschappelijk onderwijs. A dutch language slidekit used in a workshop for high school students attending the science fair organised by the Leiden University.
A.I. versus human-interfaced literature searching in medicine A slidekit used in a presentation on the developments in AI assisted literature search. The platform Evidencehunt is used as a leading example.
LLMs, academic honesty and the validity of exams. A video for the annual meet up for all exam board of the Leiden University. It includes a call to action to take a proactive stance on ensuring the validity of the examination, with a focus on examining throughout the process a student goes through, not just examine the end product. Here are the slides as pdf.
LLMs, scientific honesty and the future of scientific writing The slide kit used in the presentation for the Open Science Community Leiden with a focus on the future of academic writing.
LLMs in PhD dissertations @ LUMC – an information leaflet on how the graduate school of the Leiden University Medical Center is now monitoring LLM use in dissertations. Even when LLM use in science might be acceptable, dissertations are also a form of examination and the graduate school wants to monitor its use for potential future policy development.
Publication – ChatGPT as an author of academic papers is wrong and highlights the concepts of accountability and contributorship. An editorial written by our team in response to a journal publishing a piece which listed ChatGPT as an author.
GPT3 and academic dishonesty – discussion document. Slide kit shared with administrators at the LUMC to help kickstart/shape the discussion in various meetings/committees. The information might be a bit outdated, but the concepts still hold up.
updates on LLM via LUMC intranet – including a warning for GDPR violations and additional guidance for researchers. Links only available for LUMC employees.
Quality & Integrity of biomedical sciences 