Reality Check: Exposure to Hyper-Realistic AI-Generated Content and Its Relationship on Media Literacy Among Computer Engineering Students of Bulacan State University

Abstract

Artificial Intelligence (AI) has been developing in several ways that can generate new content, which has an impact on the media industry and the community. This study uses a quantitative research approach, which aims to determine the relationship between levels of exposure to hyper-realistic AI-generated content and levels of media literacy of computer engineering students at Bulacan State University. A total of 40 respondents answered the validated online questionnaire, containing the demographic profile, assessment of AI exposure, and an evaluation of media literacy using Likert scales and a human-AI identification test. The collected data were gathered using frequency distribution, weighted mean, standard deviation, and Spearman’s rho correlation at a 0.05 significance level. The results revealed that students often encountered hyper-realistic AI-generated content (overall weighted mean = 4.07). The respondents reported that they often encountered the diversion of AI formats (overall weighted mean = 4.12). The students highly understand the concept of AI-generated content and how to use it in other ways (overall weighted mean = 4.66). They also have a high sense of digital responsibility in determining the authenticated content across social media platforms (overall weighted mean = 4.54). Spearman’s rho indicated a weak relationship (ρ = 0.299) between exposure to hyper-realistic AI-generated content and media literacy. Their relationship is not statistically significant (ρ = 0.061). This research concluded that the digital experiences of the respondents are part of their everyday lives, revealing that they are frequently exposed to AI-generated images, videos, and music. This also fosters the media analytical and technical skills development. The findings revealed that the computer engineering students have a high level of media literacy as they can recognize algorithmic patterns and validate sources effectively.