This study examines the emergence of a spiral of anxiety among social media users on X (formerly Twitter) during discourse surrounding the 2024 Indonesian presidential election. Drawing on the spiral of silence theory, this study investigates how users respond emotionally when their political opinions diverge from the perceived majority, particularly under conditions of intense polarization. Using a quantitative sentiment analysis approach, the study analyzes 1,060 tweets collected through six election-related hashtags between November 13, 2023, and February 13, 2024. The data were classified using the Naive Bayes Classifier (NBC) method and validated using accuracy testing/confusion matrix in Google Colab. The findings show that while polarized users often express strong sentiments, a significant proportion of users adopt neutral tones, which can be interpreted as strategic silence or emotional withdrawal caused by fear of backlash. This silence does not reflect indifference but socio-political anxiety, indicating a shift from a spiral of silence to a spiral of anxiety. As a result, the quality of discourse declines, moving from fact-based debate to emotionally charged exchanges of opinion. This phenomenon contributes to unproductive political engagement and increased social fragmentation. The study offers theoretical and empirical insights into how digital platforms shape political communication in highly polarized societies.