AI vs human generated content: the interactive effects of message source, message orientation and AI literacy on perceived message effectiveness

This study examines how message source (AI-generated vs. human-generated) and message orientation (emotional vs. rational) interact to influence perceived message effectiveness (PME), and whether these relationships are moderated by AI literacy. A 2 × 2 × 2 between-subjects experimental design was employed, manipulating message source, message orientation and AI literacy level. AI literacy was assessed using the MAILS scale and the scores were used to assign participants to two groups. Participants then viewed advertisements for a fictitious thrift brand, ‘EcoFinds’, and completed measures of PME, and hypotheses were tested using three-way ANOVA. Human-generated advertisements were perceived as significantly more effective than AI-generated advertisements. A significant interaction emerged between message orientation and source type: emotional appeals were more effective when attributed to human creators, whereas rational appeals were relatively more effective when attributed to AI creators among participants with high AI literacy. AI literacy significantly moderated these congruence effects. This study extends research on AI-mediated persuasion by demonstrating that the effectiveness of AI-generated advertising depends not only on message appeal but also on source-message congruence and consumers’ AI literacy. The findings provide one of the earliest empirical examinations of these relationships in the Indian advertising context.