Dark Data - When your Data is lost and you don't even know

This is the slide deck of the talk "Dark Data - when your data is lost and you don't even know", given by Dr.-Ing. Björn Schembera (University of Stuttgart) on 1 July 2026 as part of the ELSA Webinar Series on Research Ethics, organized by the ELSA Working Group Ethics. Dark data refers to all data that is, among other characteristics, undocumented, unavailable, hidden, unstructured, biased, or erroneous. Leading analysts estimate that up to 80% of all data worldwide can be considered dark. Dark data poses economic, environmental, epistemic, ethical, and legal challenges for data management and the reproducibility of research data. This presentation introduces the characteristics and implications of dark data in science, with a particular focus on ethics. The ELSA Webinar Series on Research Ethics addresses a wide range of topics and questions related to research ethics, aiming to engage researchers, research data professionals, ethics committee members, and the general public alike. Talks are given in English and combine a 30–40 minute presentation with a 20–30 minute discussion.