Symbols significantly shape online discourse and social media dynamics, yet their impact on algorithmic content recommendations is not well understood. This study investigates how social, cultural, and political symbols influence YouTube’s recommendation system, using the 2024 Taiwanese Presidential Election as a case study. By categorizing videos into “symbol” and “no symbol” groups, we employ social network analysis, statistical measurements, and Generative AI to explore their propagation, relevance, and engagement. The results reveal that “symbol” content forms more cohesive communities and maintains a closer connection to the original topic, while “no symbol” content, despite broader appeal, experiences more topic drift. Initial engagement is slightly higher for “symbol” content, but “no symbol” content gains greater visibility over time, appealing to a more diverse audience. This research addresses the gap in understanding how symbols impact algorithmic pathways, potentially creating biases in content visibility and diversity. It provides new insights into the hidden mechanisms of social media algorithms, emphasizing the need for equitable content distribution.