Keeping Ethical and Legal Issues in Mind in Data Science Projects

Karlsruhe, November 09, 2026 — Am I allowed to use this data? What risks does an AI model pose? And how can bias or discrimination be avoided?

Anyone developing data science and AI projects must keep an eye on ethical, legal, and societal aspects in addition to technical issues. The “Ethical and Legal Guidelines for Data Scientists,” developed as part of NFDI4DataScience, provide guidance. One of  the contributors to the guideline is Stephanie von Maltzan from FIZ Karlsruhe’s Intellectual Property Rights (IGR) unit. 

From pracitioners for practicioners

To identify the challenges that researchers and practitioners actually face in data science and AI projects, interviews were first conducted with individuals from the research and applied fields. The analysis formed the basis for the guidelines. They address typical issues, such as the reliability and accuracy of data and AI applications, copyright and data protection, as well as potential biases and discrimination.

The guidelines are intended to help data scientists identify such issues early on and incorporate them into their daily work. They are explicitly not a substitute for specialized ELSA expertise. Rather, they are intended to establish a common language that facilitates communication and collaboration with experts from the respective fields.

The guidelines are supplemented by a checklist of questions covering three project phases: before, during, and after implementation. It helps data scientists reflect on potential problem areas but does not replace a legal review or expert advice. Additional materials on the individual topics round out the publication.

The “Ethical and Legal Guidelines for Data Scientists” were published on Zenodo on September 29, 2026:  Ethical and Legal Guidelines for Data Scientists