AI vs. human authorship in paintings: Differential impacts on consumers' perceived value Guangdong Provincial Graduate Education Innovation Program Project: Research on the Evaluation Index System for Practical Innovation Ability of Professional Degree Graduate Students (2024JGXM-076)

Abstract

The proliferation of artificial intelligence (AI) art generators has introduced algorithmically created paintings into mainstream consumer markets, yet consumers consistently assign lower value to AI-generated works than to human-created equivalents. This paper proposes perceived creative authenticity---consumers' subjective perception that an artwork reflects a genuine creative process characterized by human-centered values of originality, craftsmanship, and uniqueness---as a theoretically distinct mechanism explaining this value differential. Conceptually differentiated from perceived creativity, which concerns the novelty of the artistic output, perceived creative authenticity addresses the genuineness of the creator-work relationship. Drawing on authenticity theory and Psychological Ownership Theory, we develop a moderated mediation framework hypothesizing that (H1) human authorship elicits higher perceived value than AI authorship, (H2) perceived creative authenticity mediates this effect, while the indirect effects through perceived creativity, perceived effort, and anthropocentric creativity beliefs are not significant, establishing perceived creative authenticity as the primary explanatory mechanism within the tested model, and (H3) non-fungible token (NFT) authentication attenuates the mediated pathway by providing verifiable creative provenance. Across five studies, results support this framework: Study 1 conceptually replicates the human-AI value gap among Chinese consumers; Studies 2a—2c demonstrate the mediating role of perceived creative authenticity: Study 2a establishes the basic mediation pathway; Study 2b shows through parallel mediation that perceived creative authenticity uniquely mediates the effect while perceived creativity, perceived effort, and anthropocentric creativity beliefs do not exhibit significant indirect effects; and Study 2c provides causal evidence through a moderation-of-process design; and Study 3 shows that NFT authentication narrows the authenticity gap between human and AI art. These findings advance the literature on human-computer interaction by identifying process-based authenticity judgments as a key psychological mechanism shaping consumer responses to algorithmic creative products, with implications for artists and platforms navigating digital art markets.