Data & AI Ethics

Workshop Themes and Questions

Some of the themes and questions – big and small – that we hope to explore in the workshop through talks and in discussions include: 

1. Inter- and transdisciplinary bridges: Which bridges need to be built? Which shared concepts, shared language, and methods are missing? What can inter- and transdisciplinary exchange in teaching, ethical research, and product design look like? How can the different cultures of learning from data engage with each other and with data ethics? Do we need different ethics for different cultures/data-related practices? Do we need a different set of ethics for prediction-focused individual-level machine learning applications compared to classic, population-level statistical inference? How do, for instance, the philosophy of digitality, the historical and social contingencies of technology and knowledge at large, and the sociology of measurement and quantification inform data ethics? How can the ethics practices of social and cultural anthropologists (e.g., reflexivity and positionality) inform data work? How can AI ethics or technology ethics at large inform data ethics – or is there a ‘data (ethics) gap’ in these two fields which data ethicists can help to close? What can your discipline teach statisticians and data scientists about data ethics that they either do not know or do not heed yet? What do statisticians and data scientists do well in data ethics?

2. Embedded ethics and responsible data (science) workflows: How can the different disciplines that engage with data embed ethics in all stages? How do we implement this in teaching? What does it mean for a workflow to be responsible? To whom or to what does someone engaging with data owe responsibility? How does a data scientist’s responsibility relate to their workflow?

3. Power, and ethical and responsible governance of research, data, and AI: How can institutional aspects of power be addressed and what alternatives to current forms of access and governance of data and AI exist? Who and which structures should protect ethics in practice? How do we deal with gaps between data (producers) on one side and data users, data-based or even (semi-)automated decision-making on the other side? How do data and data uses shape ethics, laws, and political decision-making? How do funding and investment, as well as geolocation and ownership, affect data ethics?

4. (Ethics of) Science and its implications for data ethics: What is the current situation regarding the integrity of science and research? Are the current guidelines and principles, advocates and  authorities, forms of communication, and dissemination still appropriate, or do we need to rethink and recalibrate them in light of new challenges?

5. Consequences: How do we map the harms and benefits? If data ethics only becomes manifest in a specific context, what “general” lessons can we offer – for instance, in teaching and for abstract (theoretical) research?

6. Epistemic and ontological perspectives on data: How does the datafication of science affect our perception of the world? Which questions can or should reasonably be answered with the kind of data that is usually used - and which not? What kind of data do we use and what are the consequences of these decisions? What counts as data? Are personal data just numbers, or do they carry part of a person’s identity? How do we address multiple truths in data – and methodological multiplicity? What are the commonalities and differences between (using) data for science (i.e. knowledge creation) versus for decision-making?

7. Implications for non-applied, methodological scientists: The research of theoretical scientists (who develop and evaluate new methods and new ‘theories’) differs from that of applied scientists (who, e.g., analyze data). Yet, even theoretical researchers in, for instance, statistics, data science, and AI, cannot retreat to the position that data ethics is none of their concern. To what extent are ethical questions and implications for them different, in both research and education? Which unique perspectives and tools can theoretical scientists bring to data ethics? What are theoretical scientists missing? How can we embed ethics in the research and teaching in the more theoretical fields?

8. Current issues: What are open, important, and pressing questions in data ethics – in either your field or in general – to which solutions are not apparent, but require dedicated work, likely from multiple disciplines?

9. Defining Data Ethics. How do you define data ethics – for instance, in your field? What’s in, what’s out? What else – e.g., what other kinds of ethics – is intricately linked with data ethics?