This research critically analyzes the role of TikTok's algorithm in forming echo chambers, strengthening political polarization, and driving the virality of political content during Indonesia's 2024 Presidential Election. As a social media platform with over 112 million active users in Indonesia, predominantly aged 18-24, TikTok has become a strategic arena for digital political discourse competition. Using an interdisciplinary approach combining media studies, political science, and digital technology, this research reveals how TikTok's machine learning-based algorithm compiles individual user timelines based on engagement patterns rather than source credibility or chronology, thereby creating echo chambers that limit the diversity of political information. Findings indicate that political content with emotional elements, dramatization, and entertainment (politainment) has a higher probability of virality compared to factual content, posing serious implications for the quality of public deliberation and digital democracy in Indonesia. Based on the analysis of virality dynamics and digital polarization, this research proposes an Algorithm Democratization Model that combines aspects of transparency, regulation, literacy, platform architecture, and research to create a healthier digital ecosystem that supports democratic discourse.Keywords: TikTok algorithm, echo chamber, political polarization, virality, digital democracy