Paper Number
ECIS2026-2250
Paper Type
CRP
Abstract
This study examines how emotional congruence and the human-likeness of social media profiles influence the believability of fake news headlines. Specifically, it investigates how users evaluate headlines when they are (i) shared by Human, AI, or Anthropomorphized AI profiles, and (ii) when the emotional tone expressed by those profiles is either congruent or incongruent with the users’ own emotional state. We manipulate these factors in an online experiment using a Nordic population sample and a balanced set of two politically left- and two right-leaning fake news headlines sourced from Nordic fact-checking websites. The findings indicate that fake news is perceived as significantly less believable when it is shared by anthropomorphized AI profiles, and significantly more believable when it is shared by a profile that displays emotion which is congruent with the participant’s induced mood. These results carry important implications, as they highlight the growing threat from malicious misinformation-spreading actors who can deploy human-like bots capable of mirroring users’ emotional states.
Recommended Citation
Venade, Leonor Fertuzinhos de Matos and Gleasure, Rob, "Do You Feel What I Feel? How Profile Human-Likeness and Emotion Congruence Influence Fake News Believability Online." (2026). ECIS 2026 Proceedings. 24.
https://aisel.aisnet.org/ecis2026/cog_hbis/cog_hbis/24
Do You Feel What I Feel? How Profile Human-Likeness and Emotion Congruence Influence Fake News Believability Online.
This study examines how emotional congruence and the human-likeness of social media profiles influence the believability of fake news headlines. Specifically, it investigates how users evaluate headlines when they are (i) shared by Human, AI, or Anthropomorphized AI profiles, and (ii) when the emotional tone expressed by those profiles is either congruent or incongruent with the users’ own emotional state. We manipulate these factors in an online experiment using a Nordic population sample and a balanced set of two politically left- and two right-leaning fake news headlines sourced from Nordic fact-checking websites. The findings indicate that fake news is perceived as significantly less believable when it is shared by anthropomorphized AI profiles, and significantly more believable when it is shared by a profile that displays emotion which is congruent with the participant’s induced mood. These results carry important implications, as they highlight the growing threat from malicious misinformation-spreading actors who can deploy human-like bots capable of mirroring users’ emotional states.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.