Social Versus Emotional Loneliness: Diverging Consumer Responses to Anthropomorphic AI Service Robots

ABSTRACT Anthropomorphic design strategies are widely employed in AI service robots to deliver social benefits, such as alleviating loneliness. However, prior research has yielded mixed findings regarding how loneliness shapes consumer responses to anthropomorphic AI service robots. Across four studies, we show that these mixed findings can be reconciled by distinguishing between social and emotional loneliness. Emotionally lonely consumers respond less favorably to anthropomorphic AI service robots than socially lonely consumers. This effect is mediated by rumination. Because emotional loneliness reflects an intimacy‐domain self‐discrepancy, anthropomorphic robots that resemble human companions can make unmet intimacy needs more salient and trigger rumination about the absence of close companionship, thereby leading to less favorable responses. Furthermore, robot role framing serves as an important boundary condition of this effect: it emerges when the robot is framed as a companion but is attenuated when the robot is framed as an assistant. These findings advance consumer–service robot interaction and compensatory consumption research by demonstrating that loneliness subtype shapes how consumers interpret anthropomorphic cues, with rumination explaining why the same robot design can produce divergent responses.