As virtual influencers become common in social media advertising, this study examines how AI disclosure shapes audience trust and willingness to engage among U.S. Gen Z consumers. Drawing on the Persuasion Knowledge Model, this study examines responses to AI disclosure of virtual influencers among U.S. Gen Z consumers (N = 364). Participants were randomly assigned to view an Instagram-style post containing no, subtle, or explicit AI disclosure. Trust and willingness to engage were measured after participants first viewed the post and again after they were informed that the influencer was artificially generated and they viewed the same post again. Results showed that trust and willingness to engage declined following the universal disclosure, with larger declines among participants who initially perceived the influencer as human. Participants who initially perceived the influencer as virtual also showed a significant decline. Those exposed to an initial AI disclosure reported lower trust and willingness to engage than those who received no initial disclosure; however, subtle and explicit disclosure did not meaningfully differ. Trust was also strongly associated with willingness to engage after disclosure. The findings suggest that correcting an initial belief that the influencer was human may not fully explain disclosure responses. Increased salience of existing knowledge about an artificial source is a possible explanation.
Keywords: AI disclosure; virtual influencers; trust; persuasion knowledge; influencer advertising