HUMAN PERCEPTION OF ARTISTIC VALUE IN AI-ASSISTED 2D ANIMATION

Abstract

This study examines how an art-oriented audience evaluates artistic value and authenticity in 2D animated short videos created through different degrees of human and AI involvement. Four short videos based on the same original drawing of a kitten being petted were shown to participants: a fully human-made animation, an AI-assisted animation with human tracing and selected keyframe intervention, an AI-animated video with human retouching, and a fully AI-generated video with minimal human effort. The aim was to identify which criteria viewers use when judging creative works and whether perceived AI involvement affects their evaluation. A qualitative-dominant method was used. Participants were interviewed about their attitudes toward AI in art, shown the four videos, and asked to evaluate each work's artistic value, authenticity, emotional effect, and perceived authorship. The results suggest a consistent bias against works perceived as AI-made. Participants often claimed to value human-made art more and devalued videos they associated with AI. However, they were not always able to identify actual AI involvement, especially when AI-generated motion was traced or visually integrated by a human. The AI-assisted video was often perceived as human-made and highly valued. Viewers relied mainly on visible artifacts, line quality, perceived mistakes, emotional readability, and assumed authorship. These findings suggest that artistic value is shaped less by actual production method than by perceived human presence and the visibility of AI traces.