User Interviews: Ethical, Legal and Social Aspects

The legal and ethical implications of DS and AI are becoming an increasingly important issue for Data Scientists, Policymakers, and the General Public.

Part of our role at NFDI4DataScience is to identify these challenges and make recommendations on how to address them.

To achieve this goal, we gathered requirements through interviews with data science researchers and practitioners as well as any other domain experts involved in Data Science/AI projects. Between November 2022 and February 2024, we conducted 30 interviews with the corresponding number of data scientists. The interviewees came from a variety of backgrounds; they were working for academic institutions (universities and research centres) and the industry in a variety of roles, ranging from programmers to technical and business managers and research project primary investigators. They were based in a number of European countries and in a wide range of application domains, including healthcare, banking and financial services, humanities and social sciences.

The interviews were conducted by Maria Christoforaki, a Computer scientist and Stephanie von Maltzan, a legal scholar.

All interviews followed a semi-structured approach. Every interview was held in English and lasted approximately one hour. It was conducted online, recorded and transcribed.

Participants were asked about a self chosen data science project they had worked on, with a focus on the legal and ethical challenges they faced during the project.

Questions included, but were not limited to:

• A general description of the project and its workflow
• Ethical and legal prerequisites during the project workflow, including
  • Data provenance and related legal issues (e.g. consent, personal data and copyright concerns through scraping etc.)
  • Bias in the collection and preprocessing phase
  • Model bias and transparency
  • Technical and organisational measures
  • Specific application domain related problems (e.g. health care, engineering, humanities, etc)
• The participant’s assessment of the way they found solutions
• What they would do differently now
• If training was available: What issues would they wish to have been better informed about or received some kind of training on?

Presently, we are in the phase of correcting and anonymizing the transcripts and starting to process the interview content. After processing, editing and proofreading, the anonymous transcripts will be published and conclusions will be derived.

Call for participation

The call has ended. Many thanks to all interviewees.

Ethical, legal and social implications of Data Science and AI are turning out to be an important issue that is increasingly relevant for data scientists, policymakers and the general public.

NFDI4DataScience plans to identify these challenges and issue recommendations on how they should be addressed. In order to achieve this goal we are gathering requirements via interviews with researchers and practitioners that are/were involved in Data Science and AI projects.

Are you a Data Scientist? Are you concerned with the sometimes disparate impact of your work? Are you feeling lost in the jungle of numerous legal provisions? Then you can help us by describing the specific ethical, legal and social challenges you have faced in any of your projects. An interview will take about 60 minutes and will be conducted online.

Help us shape the future! To participate, please contact Maria Christoforaki (maria.christoforaki@uk-koeln.de) and Stephanie von Maltzan (stephanie.maltzan@fiz-karlsruhe.de).