A systematic review on the psychological implications of AI-generated visual content for user's Self-esteem and Self-perception

Abstract

Over the past decade, the use of AI-generated visual images has grown exponentially, posing novel psychological challenges. This is particularly evident in constructs such as self-esteem and self-perception, which together influence basic psychological processes underlying identity formation, social comparison, and self-worth. In contrast to traditional media, AI-generated content provides algorithmically optimized, hyper-realistic depictions that could skew standards of self-image. This systematic review examines the relation between exposure to AI- AI-generated visual information and psychological effects spanning across diverse demographicsand contexts. To identify empirical and theoretical evidence on these psychological effects of AI-generated images, a comprehensive literature search was conducted across several databases, including Scopus, PsycINFO, PubMed, Springer, ACM, and ResearchGate. Research findings show that self-esteem is affected by different AI-generated images, leading to increased social comparison driven by algorithmically refined visuals and reduced authenticity boundaries that create skewed benchmarks. Studies also show that adolescents and populations within impoverished countries are more vulnerable to this phenomenon. Gender based inequalities are especially evident when it comes to non-consensual intimate imagery and body image issues. Existing research suggests that AI-generated content may be linked to risk factors for different eating disorders and body dysmorphia. The findings underscore the need for digital literacy initiatives, ethical frameworks, and culturally aware mental health services to address the specific challenges posed by generative AI. Future research should place greater emphasis on longitudinal approaches, the development of protective mechanisms, and strategies that minimise harm while thoughtfully utilizing the positive applications of generative AI.