Social media platforms prioritize sensational or trending content, often overshadowing less popular but important topics and hindering discourse diversification. They evolve into echo chambers, where users predominantly encounter views aligned with their own. Security threat awareness for emerging technologies remains restricted, primarily because of the overconcentration of discussions influenced by both human and algorithmic factors. We seek to identify security threats related to emerging technology that are overshadowed and underrepresented due to the overconcentration of others. Next, we study uncertainty reduction approaches and emotional appraisal dimensions to understand how they contribute to the amplification or overconcentration of specific security threats. By combining computational NLP techniques to detect overconcentrated topics with scenario-based factorial surveys, this study proposes to provide a thorough examination of threat amplification in the realm of social media.