What are the gaps in accountability between algorithm projects and the society impacted by algorithms?

Gloria Miller will be presenting The Deficits in Stakeholder Accountability for Artificial Intelligence Projects at the European Group for Organizational Studies (EGOS ) on Thursday, July 08, 16:00 to 18:00.

This study uses a stakeholder accountability model and network analysis to address a question regarding the gaps in accountability between artificial intelligence projects and society. Individuals and society are considerably impacted by the outcomes and decisions that result from algorithms, but have limited or no role in the projects in which these algorithms are implemented. Furthermore, the organizations that develop, operate, and host the algorithms prefer self-regulation over new regulations or legislation. Using a mixed-method approach, this study addresses the following research question: What are the gaps in accountability between algorithm projects and the society impacted by algorithms? The findings identified six accountability, knowledge, and governance gaps. The discussion explains why the deficits are regrettable and identifies options for addressing the gaps. This information can be useful for project managers, project sponsors, and policy advisors interested in algorithmic decision-making and development processes.