Authorship disclosure and consumer perception of AI-generated graphic designEkart, Barbara (Avtor) Ahtik, Jure (Avtor) artificial intelligencegraphic designphotographyconsumer perceptionadvertisingauthorship disclosureThe increasing integration of artificial intelligence (AI) into graphic design and advertising has raised pressing questions about the role of authorship, trust and aesthetic judgement. This study examines how consumers perceive AI-generated versus human-created advertising visuals in the context of jewellery advertising. Two online surveys (n = 127) were conducted to compare participants' preferences under the conditions of disclosed and undisclosed authorship. The results show that while AI-generated visuals were sometimes rated favourably when authorship was hidden, human-created content was clearly preferred overall – especially when authorship was disclosed. A gender analysis revealed that female participants were especially sensitive to authorship cues, favouring human-created visuals. Logistic regression further confirmed that authorship disclosure, gender and design features such as human presence and serif typography were significant predictors of preference. Qualitative responses suggest that while AI visuals are technically competent, they lack emotional authenticity and narrative resonance. These findings emphasise the importance of transparency, emotional design and collaboration between humans and AI in visual communication. The study contributes to ongoing debates about machine creativity, aesthetic value and ethical disclosure, and offers practical implications for designers and marketers using AI in emotionally-driven contexts. 20262026-06-11 04:09:42Članek v reviji183303UDK: 655ISSN pri članku: 2217-379XCOBISS_ID: 281115651sl