When Influencers Are Wrong: Conditions of Blame and Trust for AI Versus Human Endorsers

Date of Award

Spring 5-2026

Degree Name

Bachelor of Science (BS)

Advisor(s)

Hoori Rafieian Koopaei

Abstract

This study examines how consumers evaluate human vs AI influencers in social media product promotion contexts, focusing on perceived quality, trust, credibility, effectiveness, and credit and blame attribution. Using a multi-study design, our first study surveyed 100 participants to compare human and AI influencers across variable judgments and anticipated outcomes for informative and emotional campaigns. The randomized second study experimentally tested whether and how influencer type shaped brand credit and blame when a product succeeded or failed, respectively. Results showed a consistent preference for human influencers. Products promoted by human influencers were viewed as higher in quality, more trustworthy, more credible, and more effective, especially for more emotional content. In attribution decisions, brand credit was similar for humans and AI influencers when a product succeeded, but brands were often blamed more when the product failure filled AI influencer promotion. These findings suggest that AI influencers may be less effective than human influencers in building relational value and may increase a brand's vulnerability in failure contexts. The study contributes to research on influencer marketing, algorithm aversion, and consumer responsibility judgements in AI-mediated communication and endorsements.

Link to Full Text

Share

COinS