Evidence for enhanced bias-control when using AI recommendations
Abstract
Decision makers’ utilization of artificial intelligence (AI) based advice, compared to other advising sources and mechanisms, has raised much attention. In this ongoing study we use the theoretical lens of motivated reasoning to analyze whether the provision of AI advice as compared to a human expert advice in a recommendation system can affect bias-control by the decision maker. Drawing on literature on group decision making, we consider embedded AI and expert advice as coming from group members in a decision-making setting. We suggest that the highly advocated peculiarity of AI-based advice of being potentially systematically biased results in better control for potential biases, both AI and the decision-maker’s own biases. Initial results from an online experiment support our hypotheses. The results of this research shed light on decision making with AI and have wide implications for design of bias-reducing information systems.
Recommended Citation
Rubin, Eran and Benbasat, Izak, "Evidence for enhanced bias-control when using AI recommendations" (2025). SIGHCI 2024 Proceedings. 12.
https://aisel.aisnet.org/sighci2024/12
