Experiencing Political Virality: A Phenomenological Study of Deepfakes, Algorithmic Mediation, and Digital Power During the National Political Protest in Jakarta, Indonesia, on 25 August 2025

Artificial intelligence (AI)-driven digital platforms increasingly shape political experiences by mediating public attention, participation, and perceptions of legitimacy. During the National Political Protest in Jakarta on 25 August 2025, AI-generated deepfake content transformed how citizens interpreted political information and negotiated digital power. Although previous studies have examined deepfakes, political virality, algorithmic mediation, and digital power, these phenomena have largely been investigated separately, limiting understanding of their integrated influence on contemporary political experiences. This study employed a digital phenomenological approach, drawing on theories of affective publics, platform power, and algorithmic mediation. Data were collected through in-depth interviews with 35 participants, digital observation, and media artifact analysis. Findings reveal that digital political experiences emerge through interactions among synthetic content, epistemic uncertainty, algorithmic amplification, affective engagement, political mobilization, and the construction of algorithmically mediated political reality. Based on these findings, the study proposes the Political Virality Ecosystem Model (PVEM), conceptualizing political virality as a cyclical communicicative ecosystem linking synthetic content production, algorithmic amplification, affective engagement, political mobilization, political reality construction, and information verification within a viral feedback loop. The model advances digital political communication theory by explaining how AI, algorithmic power, affective publics, and digital citizenship collectively shape contemporary political experiences.