How good are humans at detecting AI-generated images? Learnings from an experiment

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Roca, Thomas, Roman, Anthony Cintron, Vega, Jehú Torres, Duarte, Marcelo, Wang, Pengce, White, Kevin, Misra, Amit, Ferres, Juan Lavista
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915408524607488
author Roca, Thomas
Roman, Anthony Cintron
Vega, Jehú Torres
Duarte, Marcelo
Wang, Pengce
White, Kevin
Misra, Amit
Ferres, Juan Lavista
author_facet Roca, Thomas
Roman, Anthony Cintron
Vega, Jehú Torres
Duarte, Marcelo
Wang, Pengce
White, Kevin
Misra, Amit
Ferres, Juan Lavista
contents As AI-powered image generation improves, a key question is how well human beings can differentiate between "real" and AI-generated or modified images. Using data collected from the online game "Real or Not Quiz.", this study investigates how effectively people can distinguish AI-generated images from real ones. Participants viewed a randomized set of real and AI-generated images, aiming to identify their authenticity. Analysis of approximately 287,000 image evaluations by over 12,500 global participants revealed an overall success rate of only 62\%, indicating a modest ability, slightly above chance. Participants were most accurate with human portraits but struggled significantly with natural and urban landscapes. These results highlight the inherent challenge humans face in distinguishing AI-generated visual content, particularly images without obvious artifacts or stylistic cues. This study stresses the need for transparency tools, such as watermarks and robust AI detection tools to mitigate the risks of misinformation arising from AI-generated content
format Preprint
id arxiv_https___arxiv_org_abs_2507_18640
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle How good are humans at detecting AI-generated images? Learnings from an experiment
Roca, Thomas
Roman, Anthony Cintron
Vega, Jehú Torres
Duarte, Marcelo
Wang, Pengce
White, Kevin
Misra, Amit
Ferres, Juan Lavista
Human-Computer Interaction
Artificial Intelligence
Computer Vision and Pattern Recognition
As AI-powered image generation improves, a key question is how well human beings can differentiate between "real" and AI-generated or modified images. Using data collected from the online game "Real or Not Quiz.", this study investigates how effectively people can distinguish AI-generated images from real ones. Participants viewed a randomized set of real and AI-generated images, aiming to identify their authenticity. Analysis of approximately 287,000 image evaluations by over 12,500 global participants revealed an overall success rate of only 62\%, indicating a modest ability, slightly above chance. Participants were most accurate with human portraits but struggled significantly with natural and urban landscapes. These results highlight the inherent challenge humans face in distinguishing AI-generated visual content, particularly images without obvious artifacts or stylistic cues. This study stresses the need for transparency tools, such as watermarks and robust AI detection tools to mitigate the risks of misinformation arising from AI-generated content
title How good are humans at detecting AI-generated images? Learnings from an experiment
topic Human-Computer Interaction
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2507.18640