feminist data ethics of care for machine learning: The what, why, who and how

This article conceptualises and provides a roadmap for operationalising a feminist data ethics of care framework for the subfield of artificial intelligence (‘AI’) known as ‘machine learning’. After outlining the principles and praxis that comprise our framework, and then using it to evaluate the current state of mainstream AI ethics content, we argue that this literature tends to be overly abstract and founded on a heteropatriarchal world view. We contend that because most AI ethics content fails to equitably and explicitly assign responsibility to actors in the machine learning economy, there is a risk of implicitly reinforcing the status quo of gender power relations and other substantive inequalities, which in turn contributes to the significant gap between AI ethics principles and applied AI ethics more broadly. We argue that our feminist data ethics of care framework can help to fill this gap by paying particular attention to both the ‘who’ and the ‘how’, as well as by outlining a range of methods, approaches, and best practices that societal actors can use now to make interventions into the machine learning economy. Critically, a feminist data ethics of care is unlikely to be achieved in this context, and beyond, unless all stakeholders, including women, men, non-binary and transgender people, take responsibility for this much needed work.

The Underlying Values of Data Ethics Frameworks: A Critical Analysis of Discourses and Power Structures

A multitude of ethical guidelines and codes of conduct have been released by private and public organizations during the past years. Those abstract statements serve as a response to incidents of discriminatory algorithms and systems and have been quantitatively investigated for the proclaimed principles. The current study focuses on four frameworks designed for application during the development of new technologies. The purpose is to identify values and value conflicts and consider how these are represented in relation to established discourses, practices, and attitudes in Computer and Information Ethics. This helps to understand to what extent the frameworks contribute to social change. Critical Discourse Analysis according to Fairclough is used to examine language and discourses, and review edition and publication processes. Well-established values like transparency, non-maleficence, justice, accountability, and privacy were detected whereas value conflicts were barely addressed. Interestingly, the values were more often framed by a business, and technology discourse than an ethical discourse. The results suggest a hegemonic struggle between academia and tech industry whereas power asymmetries between developers and stakeholders are reinforced. It is recommended to extend stakeholder participation from the beginning and emphasize value conflicts. This can contribute to advance the field and effectively encourage a public debate about the desired technological progress.

The Challenges of Big Data for Research Ethics Committees: A Qualitative Swiss Study

Big data trends in health research challenge the oversight mechanism of the Research Ethics Committees (RECs). The traditional standards of research quality and the mandate of RECs illuminate deficits in facing the computational complexity, methodological novelty, and limited auditability of these approaches. To better understand the challenges facing RECs, we explored the perspectives and attitudes of the members of the seven Swiss Cantonal RECs via semi-structured qualitative interviews. Our interviews reveal limited experience among REC members with the review of big data research, insufficient expertise in data science, and uncertainty about how to mitigate big data research risks. Nonetheless, RECs could strengthen their oversight by training in data science and big data ethics, complementing their role with external experts and ad hoc boards, and introducing precise shared practices.

Data Science: Trends, Perspectives, and Prospects

Data Science is one of today’s most rapidly growing academic fields and has significant implications for all conventional scientific studies. However, most of the relevant studies so far have been limited to one or several facets of Data Science from a specific application domain perspective and fail to discuss its theoretical framework. Data Science is a novel science in that its research goals, perspectives, and body of knowledge is distinct from other sciences. The core theories of Data Science are the DIKW pyramid, data-intensive scientific discovery, data science lifecycle, data wrangling or munging, big data analytics, data management and governance, data products development, and big data visualization. Six main trends characterize the recent theoretical studies on Data Science: growing significance of DataOps, the rise of citizen data scientists, enabling augmented data science, diversity of domain-specific data science, and implementing data stories as data products. The further development of Data Science should prioritize four ways to turning challenges into opportunities: accelerating theoretical studies of data science, the trade-off between explainability and performance, achieving data ethics, privacy and trust, and aligning academic curricula to industrial needs.

Data Ethics Frameworks

Zusammenfassung Zuletzt veröffentlichten viele Organisationen ethische Richtlinien, um ihre Haltung gegen Diskriminierung durch Algorithmen zu betonen. Vier dieser Frameworks werden mithilfe der kritischen Diskursanalyse untersucht. Ziel ist es, die darin vermittelten Werte und Wertkonflikte zu identifizieren. Die Ergebnisse weisen darauf hin, dass etablierte Werte aus der Computer- und Informationsethik aufgegriffen und bestehende Machtstrukturen zwischen Akteuren verstärkt werden.