Dresden 2026 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

SYTP: Symposium Tipping Points in Social and Climate Systems

SYTP 1: Tipping Points in Social and Climate Systems

SYTP 1.4: Invited Talk

Thursday, March 12, 2026, 16:45–17:15, HSZ/AUDI

Algorithmic amplification and contextual sensitivity in political information exposure — Iris Damião, Ana Vranic, Paulo Almeida, Lília Perfeito, and •Joana Gonçalves de Sá — LIP, Lisbon, Portugal

Understanding criticality and tipping points is key to analyzing how small changes can cascade into large-scale effects. In information ecosystems, these dynamics emerge in both social media and algorithmically mediated platforms, where recommendation systems interact with user behavior to produce disproportionate societal impacts.

Misinformation on platforms like Twitter (X) is often assumed to spread faster than the truth. Yet, empirical evidence varies. Analyzing three large datasets -Vosoughi News, Global Claims, and Truth Seeker- we find that cascade size distributions are very dependent on sampling choices and this sensitivity can lead to opposite results when studying information diffusion.

Beyond social media, search engines (SEs) and large language models (LLMs) increasingly shape access to information. Using a privacy-preserving bot system, we conducted synchronized searches prior to the 2024 European Parliament. Results varied by location and query type; however, SEs and LLMs consistently emphasized right-leaning entities, revealing systematic algorithmic bias.

Together, these findings illustrate that algorithmically driven platforms can reflect and amplify political biases, creating fragmented informational realities.

Keywords: Algorithmic Bias; Political content; Sampling Sensitivity; Social Media; search Engines and LLMs