Investigating Ethical Considerations of Machine Learning Adoptions Within Organizations: A Systematic Literature Review

Abstract

The aim of this manuscript is to identify ethical issues as well as develop a conceptual framework related to the adoption of Machine Learning (ML) in organizations. The following research question was identified: 1. What are the ethical issues that organizations may potentially face as they work to design and adopt Machine Learning systems in their processes? The systematic literature format was selected to serve as the foundation for this study. Kitchenham’s (2004) seminal work for systematic literature reviews functioned as the basis for this study. This manuscript contains a brief history of ML along with conceptual terminology to establish context. Data collection included Academy of Management (35 articles), JSTOR (183 articles), and Google Scholar (16,500 articles) databases. In order to convey the rigor and reproducibility behind this effort, I utilized the PRISMA methodology and documented the process. Upon completion of the PRISMA approach, twenty-eight articles were selected for the final study and examined in-depth. A Machine Learning program, Voyant, was used to conduct the preliminary examination and a manual review of the articles followed. Consequently, I propose the Ethical Considerations Landscape Model and identify five corresponding ethical approaches to function as the theoretical foundation. The supporting ethical theories include: Utilitarian, Common Good, Fairness & Justice, Rights, and Virtue. I mapped the ethical theories to the Ethical Landscape Model as an aid for managers to guide their decision-making and actions. The result of this research is the Ethical Considerations for Machine Learning Adoption Framework.