Quick Buzz Feed
The next face helping you solve a banking problem or book a flight may belong to a celebrity who's not actually there. A study co-authored by Iowa State University researcher Lingyao (Ivy) Yuan reveals that AI-generated celebrity "digital twins" can shape customer perceptions in powerful ways, making people more likely to trust a service and even forgive its mistakes. This research delves into 'Celebrity-as-a-Service,' where companies use AI-powered or human-operated digital replicas of celebrities to interact with customers, suggesting that familiar faces create a 'halo effect' that boosts trust and willingness to engage with the service.
This section explains how digital twins represent a significant evolution from traditional celebrity endorsements. Instead of merely appearing in marketing campaigns, a celebrity's digital likeness can now interact with customers in real-time, performing tasks like answering questions, solving problems, and completing transactions. The researchers, including Lingyao (Ivy) Yuan, Antino Kim, Mike Seymour, and Alan R. Dennis, conducted experiments to test if a celebrity's familiar face could influence customer reactions, even when providing the same service. They used a digital prototype resembling actor Hugh Jackman for one group and a non-celebrity digital representative for another, ensuring both provided identical assistance. The findings showed a clear 'celebrity advantage,' with participants consistently rating the digital celebrity agent more favorably on measures like capability, benevolence, and honesty. This indicates a powerful psychological 'halo effect' where positive feelings about the celebrity transfer to the service itself, fostering trust and positive expectations.
This part of the study explored customer reactions when AI-powered technology made mistakes. In a second experiment, common customer service errors, such as mishearing names or misunderstanding addresses, were introduced. Typically, such errors erode trust and increase frustration. However, participants interacting with the celebrity version of the digital representative were significantly more forgiving and less prone to anger. The positive emotional connection associated with the real-world celebrity acted as a psychological buffer, mitigating negative reactions to service errors. Furthermore, the researchers observed no evidence of 'algorithm aversion,' meaning participants were not more critical when they knew the celebrity avatar was controlled by AI rather than a human, suggesting that the celebrity's likeness effectively masked potential biases against AI.
The research has significant implications for both companies and consumers. For businesses, these findings can inform the design of more effective digital customer service experiences, potentially leading to increased customer satisfaction and loyalty. The study suggests a future where AI-powered 'digital humans' could be customized for individual customers based on age, language, or personal preferences, enhancing personalized interactions. However, Yuan emphasizes the importance of transparency and ethical guidelines as this technology advances. She stresses that customers must always be aware when they are interacting with AI, not a human. This awareness is crucial to ensure the technology offers genuine value responsibly and ethically. Ultimately, the study concludes that while a familiar face doesn't alter a service's functional performance, it profoundly impacts how people perceive and feel about using it, highlighting the growing significance of appearance in AI-driven interactions.