by Shiyu Yang*, Nicole M. Krause, Luye Bao, Mikhaila N. Calice, Todd P. Newman, Dietram A. Scheufele, Michael A. Xenos, Dominique Brossard                                             

Abstract

Using data from a nationally representative survey of U.S. adults, this study explores how trust in key actors to responsibly manage artificial intelligence (AI) develops among members of the U.S. population and how trust, along with other key factors, shapes public attitudes toward AI. Greater trust is linked to stronger support for AI, both directly and indirectly (through risk and benefit perceptions). Furthermore, the strength or direction of the link between trust and support—as well as media diets and trust—differs significantly for liberals and conservatives, suggesting that Americans are indeed beginning to process AI-related information through a political lens.