Managing AI ethics tensions : a case study at Multichoice

Loading...
Thumbnail Image

Authors

Journal Title

Journal ISSN

Volume Title

Publisher

University of Pretoria

Abstract

Interest in Artificial Intelligence (AI) has never been greater – while we’re at a likely tipping point in the adoption of AI into mainstream industry, we’re still grappling with effective ways to manage the ethical concerns that AI surfaces on a regular basis. The purpose of this study is to shed some light into the types of AI applications being developed in the media industry in South Africa, investigate how AI ethics tensions surface and are managed when building these AI applications and provide recommendations for the management of AI ethics tensions in media organisations. Data was collected from respondents on the types of AI applications being developed, as well as the nature and characteristics of these projects including roles required to staff the projects, project duration, focus and business objectives, project outcomes, technologies used, and source of technologies used. The study reviewed recent literature on AI ethics, and specifically research into the roles played by both individuals as well as the organisations they work for, in managing AI ethics considerations. Using the insights from the literature, as well as data collected during the study via a cross-sectional survey implementation, analysis was performed to determine associations between actions in the management of AI ethics tensions and the perceived outcomes and effectiveness. Several statistically significant associations, with both weak and medium effect size, were noted between the way AI ethics tensions were discovered and managed during projects, and the perceived outcomes and effectiveness of these actions. The associations noted potentially have implications for media organisations that are implementing AI solutions and are seeking to effectively manage AI ethics tensions. Based on these analyses performed, recommendations are provided to inform the creation of effective frameworks to manage AI ethics tensions at media organisations developing AI solutions. Limitations of the study and further areas of research required are also discussed.

Description

Dissertation (MA (Philosophy))--University of Pretoria, 2023.

Keywords

Artificial intelligence ethics, AI ethics tensions, AI ethics guidelines, AI ethics codes of conduct, UCTD

Sustainable Development Goals

Citation

*