Temporal Framing in Balanced News Coverage of Artificial Intelligence and Public Attitudes

ABSTRACT

This study investigates how balanced news coverage, presenting both the positive and negative aspects of artificial intelligence (AI), may lead to attitudinal changes among the public when it is paired with different temporal frames. Different construal levels associated with temporal distance and evaluations of pros and cons provide a theoretical grounding to understand these dynamics. The results of an online experiment showed that exposure to the near-future temporal framing of AI (vs. distant-future temporal framing) induced greater perceived severity of AI risks. This, in turn, decreased public support for AI development and use. The indirect effect was conditional upon message order such that the effect was only significant when the message was presented in the risk/benefit order. The findings suggest that non-refutational two-sided frames reflecting journalistic norms of balance may lead to public opinion change because the accompanying temporal frames unwittingly impact which side people put more weight on. Thus, the selection of sources and the use of time reference need to be carefully vetted by journalists when reporting controversial science and technology.

Acknowledgments

The author would like to thank Sheila T. Murphy for providing helpful feedback on the experimental stimuli and Dmitri Williams for his constructive comments on an earlier version of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author, Sukyoung Choi, upon reasonable request.

Notes

¹ Although no specific prediction was made a priori, message order was considered as another factor to rule out any possible primacy and/or recency effects (Chong & Druckman, 2013). This approach is similar to those used in several previous studies that manipulated the order of information presentation to control for order effects (e.g., Kim et al., 2008; Orbell & Hagger, 2006; Orbell & Kyriakaki, 2008).

² Prior to the main experiment, we ran a pilot study with 222 participants via Qualtrics to see whether perceived temporal distance differs across the two temporal frame conditions (as done in the main experiment), but also to explore whether temporal framing could alter perceived temporal distance. For exploratory purposes, we tested whether perceived temporal distance could serve as a mediator between temporal framing and various outcome variables such as regulatory attitudes and intention to engage in preventive behaviors. Message order was not considered in this pilot study. A mediation analysis revealed that individuals who were exposed to a near-future temporal frame (vs. distant-future temporal frame) perceived AI’s impact as temporally closer, which, in turn, led to increased support for commercial AI regulation (b = .12, SE = .05, 95% CI [.036, .258]) and greater intentions to engage in preventive behaviors (b = .18, SE = .06, 95% CI [.064, .323]).

Additional information

Funding

The work was supported by the the USC Graduate School and and Annenberg Doctoral Student Summer Research Fellowship.

Notes on contributors

Sukyoung Choi

Sukyoung Choi (MA, Seoul National University) is a doctoral candidate at the Annenberg School for Communication at the University of Southern California. She studies media psychology and emerging communication technologies. Her research examines psychological and interpersonal processes in social media and online games, and how advances in artificial intelligence (AI) impact mediated interaction. She is also interested in exploring how the public makes sense of AI and its societal/political implications.