Skip to main content

ETHNOS_APP

Home • Search • Journals • List 0

A (Mid)journey Through Reality

Assessing Accuracy, Impostor Bias, and Automation Bias in Human Detection of AI‐Generated Images

Bibliographic Data

ID23039367
AuthorsMirko Casu (0000-0001-6975-2241, University of Catania, corresponding author), Luca Guarnera (0000-0001-8315-351X, University of Catania), Ignazio Zangara (0009-0006-4984-9031, University of Catania), Pasquale Caponnetto (0000-0002-9368-5867, University of Catania), Sebastiano Battiato (0000-0001-6127-2470, University of Catania)
EditorsMona Alhasani (0000-0001-6604-8750)
Year2025
Volume2025
Issue1
Publication date2025-01-01
Peer ReviewedYes
Open AccessYes
TypeARTICLE
VenueHuman Behavior and Emerging Technologies (JOURNAL)
Journal identifiersISSN: 2578-1863 • E-ISSN: 2578-1863
PublisherWiley (PUBLISHER • GB)
DOI10.1155/hbe2/9977058
OpenAlexW4414129468
LanguageEN
References cited54

While the challenge of distinguishing AI‐generated from real images is widely acknowledged, the specific cognitive biases that systematically shape human judgment in this domain remain poorly understood. It is particularly unclear how a general awareness of AI capabilities fosters novel biases, like a pervasive skepticism (“impostor bias”), and how this interacts with established phenomena like “automation bias”. This study addresses this gap by providing the first quantitative analysis of how these two biases operate across five distinct experimental variants designed to test the context‐dependency of human perception. Through a mixed‐methods study with 746 participants, we demonstrate that human authentication accuracy hovered around chance levels (ranging from 47.0% to 55.5%). However, our analysis provides robust evidence for the systematic operation of cognitive biases. We validate the presence of “impostor bias” through a consistent pattern of higher doubt for AI‐generated images and confirm “automation bias” through significant opinion changes following algorithmic suggestions. Our findings reveal that these biases are not uniform across populations: gender was a consistent predictor of automation bias, with males in all five variants showing a significantly stronger and more consistent tendency (Cohen’s d = 0.254–0.683) to be influenced by algorithmic suggestions. In contrast, age and academic background had minimal and highly localized effects. Furthermore, we identified a significant interaction between experimental stimuli and performance over time, isolating a pronounced fatigue effect to a single questionnaire variant where accuracy progressively declined (by approximately 1.7% per trial). By integrating human feedback with Grad‐CAM visualizations, we confirm a divergence between human holistic evaluation and the localized focus of machine learning models. These findings carry direct implications for policy, as discussed within the context of the European AI Act, and inform the design of human–AI systems and media literacy programs aimed at mitigating these critical cognitive vulnerabilities.

Automation · Cognition · Divergence (linguistics) · Domain (mathematical analysis) · Human intelligence · Skepticism · Artificial Intelligence in Healthcare and Education · Ethics and Social Impacts of AI · Explainable Artificial Intelligence (XAI

  • High-Resolution Image Synthesis with Latent Diffusion Models

    Open Access•Robin Rombach, Andreas Blattmann et al.•2022 IEEE/CVF Conference on…•2022

  • Art and the science of generative AI

    Open Access•Ziv Epstein, Sudi Nurtini et al.•Science•2023

  • ImageNet Large Scale Visual Recognition Challenge

    Open Access•Olga Russakovsky, Jia Deng et al.•International Journal of Computer…•2015

  • Deepfakes as narratives

    Open Access•María T Soto-Sanfiel, Ariadna Angulo-Brunet et al.•Computers in Human Behavior•2025

  • Providing detection strategies to improve human detection of deepfakes

    Open Access•Klaire Somoray, Dan J Miller•Computers in Human Behavior•2023

  • Flow in ChatGPT-based logic learning and its influences on logic and self-efficacy in English argumentative writing

    Open Access•Ruofei Zhang, Di Zou et al.•Computers in Human Behavior•2025

  • Fake images

    Open Access•Cuihua Shen, Mona Kasra et al.•New Media & Society•2018

  • Deepfakes and Disinformation

    Open Access•C Vaccari, Andrew Chadwick•Social Media + Society•2020

Citation velocityhistorical
Highly citedNo
Ethnos_APP • Open Source Project • MIT License • Frontend v2.0.0 • Privacy and Cookies • API Documentation: api.ethnos.app/docs • API Source Code: GitHub • DOI: 10.5281/zenodo.17049435 • Frontend Source Code: GitHub • DOI: 10.5281/zenodo.17050053 • cruz.rio.br • Expectantes Misericordiae