Identifying AI Bias and Mitigation Challenges Through a Socio-Technical Perspective
Author(s)
Date Issued
2024
Type
conferenceObject
Abstract
In an era where Artificial Intelligence (AI) increasingly permeates the global landscape, its potential to impact many people necessitates heightened vigilance against biases in AI-driven outputs. This research employs a grounded methodology, gathering insights through semi-structured interviews with AI professionals engaged in various stages of the AI lifecycle. Our objective is to develop a socio-technical framework that systematically identifies and mitigates AI bias, including various aspects of stakeholder participation. Initial findings from this study have refined our interview protocols and expanded our understanding of overlooked AI biases. These insights offer a deeper understanding of the influence exerted by non-end-user stakeholders on the AI development lifecycle, providing a multi-dimensional perspective on the nuances of AI bias and its emergence and mitigation.
Conference(s)
ECIS 2024 People First: Constructing Digital Futures Together