Simultaneous pursuit of accountability for regulatory compliance, financial benefits, and societal impacts in artificial intelligence (AI) projects

This exploratory study, grounded in agency theory, employs quantitative analyses to investigate the simultaneous pursuit of accountability for regulatory compliance, financial benefits, and societal impacts within artificial intelligence (AI) projects.An agent-principal matrix was developed, synthesizing knowledge from the AI stakeholder model into 11 accountability indicators.These indicators establish a standard of responsibility among project actors for regulatory compliance, ethical practices, and financial benefits.Using quantitative methods, we analyzed survey data on accountability and defined the scope of AI systems under development.We identified two clusters of AI systemsautonomous and non-autonomous-based on seven features.We then examined how these two types of systems, as well as the importance of sustainability and fairness, impact the promotion of accountability.Results indicate that accountabilities shift based on the scope of the AI system and the project role.Regulatory compliance, financial benefits, and societal impacts are not mutually exclusive project goals and coexist.The findings quantify subjective and theoretical speculation about accountabilities within AI projects.Additionally, the study contributes empirical data to the literature on AI, ethics, and project management.