This study aims to examine the visual expression characteristics ofadvertising videos produced with generative AI technology and to analyze howthese visual elements influence audience responses. First, a theoretical reviewidentified key visual feature...
This study aims to examine the visual expression characteristics ofadvertising videos produced with generative AI technology and to analyze howthese visual elements influence audience responses. First, a theoretical reviewidentified key visual features that frequently appear in generative AIadvertisements, including realism-based representational expressions andexpressionist approaches involving composited or surreal virtual imagery.
Based on this theoretical foundation, an analysis of generative AI advertisingcases from major domestic companies and public institutions revealed thatdifferent visual expression types induce distinct cognitive and emotionalresponse patterns among viewers.
Subsequently, industry- and type-specific frequency analyses were conductedon collected AI advertisements, showing that the F&B, finance, and IT sectorsdemonstrate particularly high levels of AI advertising utilization, with a strongprevalence of realism-oriented visual expressions. Drawing from theseclassification results, this study designed an empirical framework centered onfour audience experience factors: information processing (attention,recognition, memory), emotional response (enjoyment, interest, affinity),immersion (concentration, loss of time awareness, visual intensity), andentertainment (fun, curiosity, novelty).
An experiment involving 203 participants was conducted to verify differencesin audience responses across the visual expression types. The findings showsignificant variations between types: product-realism advertisements elicitedhigher information processing and perceived trust, whereas virtual-characterand virtual-world types tended to strengthen emotional affinity, entertainmentvalue, and visual interest. Overall, the visual expression styles of generative AI advertisements were found to be key determinants shaping both cognitiveand emotional audience responses.
By systematically categorizing the visual expression types of generative AIadvertising and empirically examining audience response differences acrossthese types, this study provides meaningful implications for future AI-basedadvertising production and design strategies.