Video: Doctoral Student Yoo Jin Chung Discusses AI-Enabled Virtual Influencers on Social Media

Doctoral student Yoo Jin Chung is in her final year of doctoral program in advertising, working on a dissertation that explores social media users parasocial relationship with virtual influencers. She was interviewed on Sept. 7, 2021 about her research on AI-enabled virtual influencers on social media, and how social media users react to virtual vs. human influencers.

Below the video is an edited transcript of the full interview.

Tell us about your research on AI-enabled virtual influencers.

Yoo Jin Chung: My research is focused on a conceptual and operational definition of virtual influencers, which we often call AI influencers, in comparison with other types of endorsers on social media, such as celebrity endorsers and influencers. My research also explores how social media users build the relationship with virtual influencers on social media through a parasocial relationship. I mainly focus on the perception of virtual influencers as a parasocial persona using parasocial attributes, which is perceived humanness and perceived interactivity.

Have you formulated any hypotheses about this research?

Yoo JIn Chung: Under the big research question, how can we differentiate virtual influencers on social media with celebrity endorsers or influencers, I’m hypothesizing that perceived humanness of virtual influencers and perceived interactivity will be positively related to parasocial relationships by virtual influencers among social media users.

I also hypothesize that physical and social attractiveness of virtual influencers will be positively related with parasocial relationship experience of social media users. Also, these parasocial relationships positively influence social media users’ attitudes towards virtual influencers, purchase intention and intention to engage with social causes on social media.

What role does artificial intelligence play in this technology?

Yoo Jin Chung: Virtual influencers, thanks to the great technology these days, have physical appearances almost identical to real humans. Some virtual influencers are difficult to recognize as non-human. They are very real. So one important thing in this research is whether social media users believe the virtual influencer is indeed a human or a machine.

Just to recognize virtual influencers as a human or non-human may cause differences in how people get influenced and sometimes how people believe the messages generated by virtual influencers. These artificial intelligence aspects from physical appearances to how they generate the information on social media will influence social media users.

What are the implications of this research?

Yoo Jin Chung: First of all, I hope to gain an operationalized definition of virtual influencers compared to other human endorses on social media. Nowadays, sometimes even the distinction between actual celebrity and micro celebrity, which we often call influencers, are not clear and not operationalized enough to measure. So, in this research with the measure of perceived humanness and permissive interactivity, I would like to have a clear definition. What is measurable? Which can differentiate virtual influencers with other types of human influencers?

Also, I want to understand how social media users build relationships and becomes a fan of virtual influencers. Whether the social media users perceive virtual influencers as a machine or human can also change how these virtual influencers and the marketing data behind them generate information as well. For instance, if the engagement of social media users using that AI technology is increased, they can reply back to thousands of comments on their social media. These are the things that I want to expect to gain from this research.