Research on AI ethics tends to examine the subject through philosophical, legal, or technical perspectives, largely neglecting the sociocultural one. This literature also predominantly focuses on Europe and the United States. Addressing these gaps, this article explores how data scientists justify and explain the ethics of their algorithmic work. Based on a pragmatist social analysis, and of 60 semi-structured interviews with Israeli data scientists, we ask: how do data scientists understand, interpret, and depict algorithmic ethics? And what ideologies, discourses, and worldviews shape algorithmic ethics? Our findings point to three dominant moral logics: (1) ethics as a personal endeavor; (2) ethics as hindering progress; and (3) ethics as a commodity. We show that while data science is a nascent profession, these moral logics originate from the techno-libertarian culture of its parent profession—engineering. Finally, we discuss the potential of these moral logics to mature into a more formal, agreed-upon moral regime.
Bibliographical noteFunding Information:
We thank the reviewers for the enlightening comments and enriching dialogue. This article was written with the support of the Shapiro Fund Fellowship for postdoctoral students, the Department of Sociology and Anthropology, Tel Aviv University.
© The Author(s) 2023.
- AI ethics
- algorithmic ethics
- data science
- moral regimes
ASJC Scopus subject areas
- Sociology and Political Science