Coverage of ethics within the artificial intelligence and machine learning academic literature

• Lillywhite, Aspen; Wolbring, Gregor

Abstract

• Disabled people are often the anticipated users of scientific and technological products and processes advanced and enabled by artificial intelligence
• also impacted by societal impacts of AI/ML
• problems have been identified in how ethics discourses engage with disabled people
• Of the n = 1659 abstracts engaging with AI/ML and ethics downloaded from Scopus (which includes all Medline articles) and the 70 databases of EBSCO ALL, we found 54 relevant abstracts using the term “patient” and 11 relevant abstracts mentioning terms linked to “impair*”, “disab*” and “deaf”

Study design

• We used a modified scoping review drawing from (Arksey & O’Malley, 2005) as the most appropriate approach for the study given the aim of our study
• Scoping studies “map rapidly the key concepts underpinning a research area” (Arksey & O’Malley, 2005, p. 21), to identify the extent of research conducted on a given topic (Davis, Drey, & Gould, 2009; Grant & Booth, 2009) and the current understanding of a given topic (S. Anderson, Allen, Peckham, & Goodwin, 2008).

Data sources

• EBSCO ALL, an umbrella database that includes over 70 other databases itself and Scopus, which incorporates the full Medline database collection
• The first article with the term “ethic*” and any of the AI terms within the EBSCO databases was published in 1981, while the first article within Scopus was published in 1962

Limitation

• Our findings also do not cover all words one could use to depict disabled people and as such, our results can not be generalized to every disability term

Discussion

• Many ethics issues pertinent to disabled people were discussed within the abstracts; for example the ethical decision making of robots but without engaging with disabled people.
• However, many barriers have been identified for disabled people to shape technology governance discussions in an anticipatory way
• including that the medical imagery of disabled people is seen to hinder their involvement in policy discussions (Wolbring, Mackay, Rybchinski, & Noga, 2013)
• In one study, it is acknowledged that ethical, moral, social, cultural, and political issues have been traditionally de-emphasized in research guided by usability concerns (Fallman, 2010)
• Given the breadth of academic disciplines, including disability studies covered by the two databases, and given that we found only one article coming from a disability studies program based out of Bremen, Germany (Bruhn et al., 2006), it might be warranted to investigate how academics choose their topics of investigation and why the topics we found lacking in the literature were not chosen
• Another angle of investigation could be why students are not acting as knowledge producers on the topics we found lacking
• As to disabled students, based on a study that investigated the experience of disabled postsecondary students in postsecondary education (Hutcheon & Wolbring, 2012) we suggest that the experience reported (feeling medicalized, hesitant to self-advocate, to try to fit in with the norm) might be factors that hinder disabled students to be knowledge producers especially on contentious issues such as ethics and disabled people