Social media shape political discourse during elections, especially in geopolitically contested regions. This work examines the 2024 Moldovan presidential election by analysing publicly available posts from X, Facebook, and Reddit collected between August and October 2024. We combine multilingual stance detection, coordination analysis, and causal inference to study both descriptive and causal mechanisms of online political discourse. First, large language models classify stance toward European integration, Russia, and the two main candidates, revealing clear cross‑platform and cross‑lingual expression differences. Second, we build retweet‑based and semantic alignment networks on X to identify amplification and aligned framing of dominant narratives. Finally, a proximal causal inference framework estimates peer influence on stance formation, showing consistently positive and statistically significant average treatment effects across major stance categories. The results indicate that amplification and peer exposure reinforce political narratives online. This is the first systematic analysis of social‑media discourse surrounding the 2024 Moldovan presidential election, demonstrating the value of integrating language annotation, network analysis, and causal inference for studying political influence in multilingual, geopolitically sensitive contexts.
Review