Cognition Under Siege: Narrative Strategies and Cognitive Biases in Contemporary Disinformation

Abstract

In recent years, misinformation campaigns have evolved significantly in both scale and sophistication. With enhanced modern technologies, malicious actors are equipped with a broader repertoire of tools to leverage our cognitive vulnerabilities, risking manipulation of perception [1]. For instance, during the COVID-19 pandemic, false information about evacuees from Wuhan circulating on social media incited riots in Eastern Ukraine [2]. In another case, Russian outlets disseminated narratives portraying Europe as a dictatorship, claiming that terrified citizens were experiencing severe restrictions on freedom of expression [3]. In such examples, misinformation is not merely an instrument or tool of deception but a strategic weapon in cognitive warfare, used to undermine trust in public authorities or to elicit emotional responses, and ultimately to destabilize collective understanding of a situation. Through the propagation of such false information, one can systematically influence an individual’s thought processes and alter their situational perception. This form of cognitive infiltration can lead to dramatic shifts of worldviews, values, knowledge, and personal or collective consciousness, often in service of a political or ideological agenda.

This thesis aims to investigate I) which cognitive vulnerabilities or biases are most frequently exploited in a selected range of case studies, and II) the rhetorical and visual narrative framing strategies used to target these vulnerabilities. The primary dataset will be derived from “EUvsDisinfo”, a flagship initiative of the European Union that monitors and debunks disinformation. The research methodology will follow a four-phase process: First, a literature review will establish clear definitions of cognitive vulnerabilities, as well as criteria for identifying them in disinformation contexts. Next, a large sample of misinformation cases from the selected database will be labeled according to the narrative strategies employed, the cognitive vulnerabilities exploited, and the underlying socio-political context. A sentiment analysis will also be used as part of this phase to assess the cognitive effects of these narratives, such as fear, confusion or distrust. After this, the annotated components will be visualized in a network diagram to uncover meaningful relationships between strategies, biases and effects. Lastly, if the scope permits, a machine learning model will be developed to predict likely narrative strategies employed by specific threat actors based on prior patterns and feature analysis. The overall purpose of this thesis is to map the psychological mechanisms behind disinformation to bolster our understanding of how belief systems and perceptions are targeted through our information environments.

References

[1] A. Sanchez-Acedo, A. Carbonell-Alcocer, M. Gértrudix & J.L. Rubio-Tamayo, “The challenges of media and information literacy in the artificial intelligence ecology: deepfakes and misinformation,” Communication & Society, pp. 223 - 239, 2024. doi: 10.15581/003.37.4.223-239

[2] EUvsDisinfo. "Protests in Novi Sanzhary are a symbol of the Maidan in Ukraine," Disinformation Case Details, 2020. [Online]. Available: https://euvsdisinfo.eu/report/protests-in-novi-sanzhary-are-a-symbol-of-the-maidan-in-ukraine/. [Accessed: May 26, 2025].

[3] EUvsDisinfo. "There is more freedom of speech in Russia than in Europe," Disinformation Case Details, 2025.  [Online]. Available: https://euvsdisinfo.eu/report/there-is-more-freedom-of-speech-in-russia-than-in-europe/. [Accessed: May 26, 2025].