AI and Deepfakes: The Future of Political Campaigns in the United States
Files
TR Number
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This paper examines how individuals engage with and view the acceptability of artificial intelligence (AI), focusing on how partisanship and disclosure affect acceptability within the context of political campaigns in the U.S. This manuscript is guided by the Persuasion Knowledge Model, Motivated Reasoning, Social Identity Theory, and Cognitive Dissonance. A 2x2 factorial experiment was conducted to test the significance of both partisanship and disclosure. The experiment also tested how AI sentiment, AI knowledge, and AI skepticism all interact to impact acceptability. Results indicated that disclosure had no significance in acceptability. Further, there was no significant relationship between partisanship and disclosure on acceptability. However, the results found a significant relationship between partisanship and the acceptability of AI usage despite skepticism. The results also showed that conservative participants were more willing to vote for a copartisan candidate and accept the use of AI to a greater extent than liberal and moderate participants were for their partisan candidate, even though they perceived themselves to be less knowledgeable about AI than liberal and moderate participants. This study contributes to the theoretical understanding of political message processing in the context of AI use and offers implications for communication strategies.