The rise of right-wing populism and increasing political polarization have amplified extremist ideologies, eroding democratic norms and fostering social conflict. Traditional interventions targeting rigid beliefs often backfire, as direct challenges provoke defensiveness and strengthen existing convictions. This study proposes an alternative approach: leveraging parasocial relationships (PSRs) with beloved public figures, such as Donald Trump, to induce cognitive dissonance by presenting contradictory information targeting peripheral beliefs. Drawing on the BENDING model, which conceptualizes belief systems as interconnected networks, the research explores how changes in peripheral beliefs can disrupt ideological rigidity. The study will use experimental methods to measure belief networks and PSRs, exposing participants to Trump statements that contradict peripheral beliefs. Analyses, including ANOVA and network modeling, will assess belief network disruptions and shifts in PSRs. This research aims to deepen understanding of belief dynamics and inform strategies to reduce extremism, mitigate polarization, and strengthen democratic resilience through subtle, non-confrontational interventions.