Authenticity in Virtual and Human Influencer Donation Appeals: Effects of Emotional Expression and Language Style

Virtual influencers (VIs) are increasingly deployed in prosocial campaigns, yet evidence regarding their effectiveness relative to human influencers (HIs) remains limited. This study examines how influencer type (VI vs. HI), emotional expression (happy vs. sad), and language style (formal vs. informal) shape perceived authenticity and subsequent donation intention. Using a 2 × 2 × 2 between-subjects experiment (N = 400, U.S. adults), we demonstrate that authenticity perceptions for HIs and VIs depend on different combinations of emotional and linguistic cues. Although HIs generally maintain an authenticity advantage, pairing happy emotion with informal language decreases perceived authenticity for HIs while enhancing perceived authenticity for VIs. Under this condition, VIs effectively reverse the typical human authenticity advantage, generating higher perceived authenticity and a favorable indirect effect on donation intention. The findings extend behavioral realism theory by showing that emotional expression and language style jointly shape authenticity perceptions of VIs while revealing distinct authenticity mechanisms for VIs and HIs.