Synthetic Politics: Prevalence, Spreaders, and Emotional Reception of AI-Generated Political Images on X (2502.11248v2)

Abstract: Despite widespread concerns about the risks of AI-generated content (AIGC) to the integrity of social media discourse, little is known about its scale and scope, the actors responsible for its dissemination online, and the user responses it elicits. In this work, we measure and characterize the prevalence, spreaders, and emotional reception of AI-generated political images. Analyzing a large-scale dataset from Twitter/X related to the 2024 U.S. Presidential Election, we find that approximately 12% of shared images are detected as AI-generated, and around 10% of users are responsible for sharing 80% of AI-generated images. AIGC superspreaders--defined as the users who not only share a high volume of AI-generated images but also receive substantial engagement through retweets--are more likely to be X Premium subscribers, have a right-leaning orientation, and exhibit automated behavior. Their profiles contain a higher proportion of AI-generated images than non-superspreaders, and some engage in extreme levels of AIGC sharing. Moreover, superspreaders' AI image tweets elicit more positive and less toxic responses than their non-AI image tweets. This study serves as one of the first steps toward understanding the role generative AI plays in shaping online socio-political environments and offers implications for platform governance.

Overview of AI-Generated Content on Social Media During the 2024 U.S. Presidential Election

This paper investigates the prevalence and dissemination of multimodal AI-generated content (AIGC) on the social media platform $\mathbb{X}$ during the 2024 U.S. Presidential Election. The authors focus on delineating the patterns and entities responsible for sharing AI-generated texts and images. Through a comprehensive dataset, the paper reveals significant insights into the concentration of AIGC and the behaviors of so-called "superspreaders" who are critical to its distribution.

Key Findings

1. Prevalence of AIGC: The analysis showed that AI-generated images vastly outnumber AI-generated texts in the observed dataset. Approximately 12% of images were AI-generated compared to about 1.4% of text content. This difference underscores the dominance of visual AIGC over textual forms in platform engagement during the election discussions.
2. Superspreaders of AIGC: A concentrated group of users, termed superspreaders, dominated the dissemination of AIGC. Notably, around 3% of text spreaders and 10% of image spreaders accounted for 80% of AIGC in their respective modalities. This pattern mirrors trends in misinformation spread, where a small group holds significant sway over content propagation.
3. User Characteristics: Superspreaders often had right-leaning political orientations, subscribed to $\mathbb{X}$ Premium, and displayed bot-like behaviors. This demographic skew suggests that AIGC dissemination is not random but correlated with specific user traits and platform engagement strategies.
4. Modal Differences: Within user profiles, AI-generated image sharers exhibited a higher proportion of AIGC compared to text sharers, indicating a stronger inclination towards visual content. Additionally, superspreaders were characterized by their automated behavior, with bot-like accounts playing a substantial role in spreading AI-generated images.

Implications

The findings of this paper offer several implications for both platform governance and future research.

• Platform Governance: The identification of superspreaders and their characteristics can inform social media platforms’ strategies to moderate AI-generated content. Governance frameworks may need to enforce stricter adherence to ethical standards for content generation and sharing, especially given the potential for such content to influence public opinion and electoral outcomes.
• Policy and Public Awareness: There is an evident need for public awareness campaigns that inform users about the nature of AIGC and its implications. By understanding the drivers behind AIGC propagation, policymakers can craft interventions that enhance digital literacy and critical evaluation of content among social media users.
• Future Research Directions: The paper sets the stage for further inquiries into how AIGC may evolve in other social contexts or platforms. Future studies could also explore the long-term effects of AIGC on user attitudes and the overall information ecosystem.

The investigation highlights a critical intersection of technology, politics, and media, emphasizing the necessity for continuous monitoring and analysis as AI technologies further integrate into socio-political discussions online. By providing a detailed overview of AIGC dynamics, this paper contributes to our understanding of AI's role in shaping modern digital interactions and societal discourse.