Data Ethics of Power – A Human Approach to Big Data and AI
Imagine an AI robot that sieves through pictures containing predominantly white faces deciding what human ‘beauty’ means. Ponder on an online search system that learns from news articles to recognize words such as ‘nanny’ and ‘receptionist’ as female and words such as ‘architect’ and ‘financier’ as male. Reflect on a mass surveillance global intelligence network or a mass political profiling campaign enabled invisibly by social networking sites with the personal data of millions of people.
These sociotechnical data systems and practices are ethically questionable. Most often they are ‘unfair’, certainly they are not morally ‘good’ but ‘bad’, and in some contexts they might even be deemed illegal. However, data ethics is not a legal assessment, neither is it a moral evaluation of the good or the bad. A ‘data ethics of power’ is concerned with making visible the power relations embedded in big data and AI sociotechnical infrastructures.
This study presents a “data ethics of power” as an action-oriented analytical framework concerned with making visible the distribution of power and power relations in the Big Data Society and the conditions of their negotiation and distribution, in order to point to design, business, policy, social and cultural processes that support a human-centric distribution of power.
The PhD thesis Data Ethics of Power – A Human Approach to Big Data and AI by Gry Hasselbalch was submitted at University of Copenhagen/Faculty of Humanities January 2021 and defended 13th of April 2021 with a committee consisting of Klaus Bruhn Jensen, Safyia Umoja Noble and Simone Van der Hof.
The conclusion and introduction can be accessed here
Interested in the PhD thesis? Please send an email to email@example.com
Data ethics is not an evaluation of the ‘good’ and the ‘bad. It’s about who gets to decide what is ‘good’ and what is ‘bad’ and ultimately what human composite is ethically responsible.
About the study
Data no longer just capture politics, economy, culture and lives. Today data is extended and ingrained in society in increasingly complex digital systems developed to contain and make sense of big amounts of it and to act on that knowledge. These digital data systems form a key component of decision-making in politics, culture, industries, and on life trajectories, and consequently they are also the center of power negotiations between different interests. As such, power is at the center of ethical concerns with data systems.
With point of departure in the author’s immersion in ‘data’ and ‘Artificial Intelligence’ policy, industry and civil society activities of the 2010s in Europe, which among others included the EU’s High-Level Group on AI and the Danish government’s first data ethics expert group, Gry Hasselbalch investigates the role of data ethics and a human (-centric) approach to the ethical governance of AI and big data.
The first two parts of the study explore the power structures for human agency and experience of what is referred to as respectively Big Data SocioTechnical infrastructures (“BDSTIs”) and Artificial Intelligence SocioTechnical Infrastructures (“AISTIs”). These are described as not only a critical concern of a ‘data ethics of power’, but of “data ethical governance” in general.
In the last part of the study, two quintessential sub-questions are posed when exploring the role and importance of human ethical agency and responsibility in the Big Data Society: Why is a ‘data ethics of power’ important? How can a ‘data ethics of power’ achieve the ‘good society’? The answers to these two questions constitute the formative framework for a ‘data ethics of power’.