ARTIFICIAL INTELLIGENCE VS. ARTIFICIAL INTELLIGENCE DEMOCRATIC TRUST: A QUANTITATIVE ANALYSIS OF POLITICAL DEEPFAKES ON SOCIAL MEDIA

Abstract

The development of artificial intelligence (AI) technology has given rise to the deepfake phenomenon, a human image synthesis technique capable of manipulating audio and video in highly realistic ways. In the political realm, this manipulation has the potential to become an instrument of disinformation that undermines the democratic order, particularly among first-time voters who consume political information massively through social media. This study aims to quantitatively analyze the impact of exposure to political deepfake content on social media on the decline in public trust among first-time voters. Using an explanatory quantitative approach, data were collected through a structured online questionnaire distributed to a sample of first-time voters aged 17–22 who actively use platforms such as TikTok, Instagram, and X (Twitter). The sampling technique used was purposive sampling with linear regression-based data analysis to test the level of significance of the influence between variables. The results show that the intensity of exposure to political deepfake content has a significant negative effect on the level of trust of first-time voters in the election process, political institutions, and the validity of digital information in general. As the level of exposure to this manipulative technology increases, the digital skepticism of information consumers also increases, which in turn erodes their trust in the pillars of democracy. The implications of this research emphasize the urgent need for strengthening artificial intelligence-based digital literacy (AI literacy) and stricter regulations for monitoring digital content to mitigate the destructive impact of synthetic disinformation on election integrity.