Impact of Deepfakes on Digital Political Disinformation in Ecuador: Analysis on TikTok

Abstract

The expansion of artificial intelligence has intensified the production of manipulated content in digital environments, particularly through deepfakes, posing new challenges for contemporary political communication. This study analyzes the impact of deepfakes on digital political disinformation in Ecuador within the TikTok platform. A mixed-methods, non-experimental design with an exploratory scope was employed, based on the analysis of 30 purposively selected posts according to criteria of audiovisual manipulation, political relevance, and user engagement. Data collection was conducted through content analysis using a structured coding matrix, with quantitative processing performed in Jamovi and qualitative analysis conducted in ATLAS.ti. Findings show that manipulated content exhibits high viral potential, particularly in video format, and is associated with elevated levels of user interaction. From a qualitative perspective, negative emotions such as distrust, indignation, and rejection dominate, alongside discursive patterns of political polarization. It is concluded that deepfakes operate not only as tools of disinformation but also as catalysts of affective engagement and fragmentation of the digital public sphere. This study provides empirical evidence on the interaction between artificial intelligence, algorithmic recommendation systems, and political behavior in social media.