Abstract
The artificial intelligence (AI) is transforming the modern branding practice with more companies using AI-driven agents, like chatbots, virtual assistants, and generative systems, to create an illusion of human-like communication with their consumers. Brands are becoming quasi-social in digital space through the tone of conversation, personalised responses and emotion, creating digital social actors within the digital space. Although the concept of anthropomorphism has been known as an effective brand strategy tool, AI opens a more dynamic and interactive mode of anthropomorphism that is adaptive, data-driven, and relational. Although AI-based anthropomorphic cues have become increasingly relevant to managers, a lack of empirical studies investigates the effect of anthropomorphic cues on consumer-brand relationships and behavioural responses.
This development paper is a nearly full-fledged mixed-methods study that explores the effects of AI-based brand anthropomorphism on perceived authenticity, brand attachment, and purchase intention in the UK digital marketplace. Based on the social response theory and authenticity theory, the study theorises interactive anthropomorphism as a unique phenomenon in the branding mediated by AI. Sequential research design was taken. First, 22 semi-structured interviews with UK consumers examined lived experiences of AI brand agents in online retail setting and service setting. The thematic analysis has shown that there are three common themes perceived warmth and relational comfort, transparency sensitivity, and authenticity tension. These results were used to design a controlled online experiment (n = 290), which controlled anthropomorphic cues (low vs. high human-likeness) and transparency disclosure (AI disclosed vs. undisclosed).
Prelude regression and mediation analyses show that moderate anthropomorphism cues have a positive effect on emotional attachment and perceived warmth that positively affect the purchase intention. Nonetheless, when the human-likeness is high, and there is no transparency, perceived authenticity decreases greatly, and scepticism is evoked. The findings of mediation indicate that an understanding of perceived authenticity and brand attachment mediates the relationship between anthropomorphic AI signals and purchase intention in sequence. Negative authenticity effects are reduced by transparency disclosure and thus transparency disclosure is a governance mechanism in AI branding.
This research has three implications to management and marketing scholarship. First, it propagates a conceptual difference between non-interactive and interactive anthropomorphism in AI situations. Second, it finds a curvilinear impact of anthropomorphism cues on authenticity perceptions. Third, it places the concept of authenticity as a key psychological intermediary between the aspects of AI design and the outcomes of relation and behavioural brands.