Title

Seeing Is No Longer Believing: Teaching Visual Literacy in the Age of AI

Creator

Kali Pikl

Rights

In Copyright.

https://rightsstatements.org/page/InC/1.0/?language=en

supported by

Ethan McMahan

Abstract

Media literacy and close analysis of multiple formats of information are both critical skills in our technology-rich environment. This is particularly the case given the rate at which Artificial Intelligence models are advancing. The current study sampled 87 university students and sought to evaluate participants’ ability to detect AI-altered/generated images. This was measured via survey and was followed by a training module designed to teach participants about some of the key artifacts in AI-generated/altered images. Participants completed a second survey following the training portion to evaluate for change. It was hypothesized that there would be a significant difference between the participant’s pre-treatment and post-treatment scores following the training module. A paired-samples t-test showed that post-treatment survey scores (M = 11.66, SD = 2.25) were significantly higher than pre-treatment scores (M = 9.05, SD = 2.61), t(86) = -10.19, p < .001, 95% CI [-3.12, -2.10]. After controlling for relevant individual differences, the results showed a large effect size (d = 1.09). This shows the ability to train audiences in how to recognize images that have been generated or altered by AI. The current study showed a slight decrease in participants’ accuracy in detecting real images due to increased skepticism. The applicability to the generalized population of the specific tips provided in this study should be studied further as image-generating AI models will likely phase out many of the artifacts used in this study.

Key Words

Artificial intelligence

media literacy

misinformation

Type

Text; Image; StillImage

Date

4/10/2026

Identifier

PURE_2026e