Artificial intelligence (AI) is an essential strategic element for many industries. Practitioners are exploring ways to incorporate AI into their products, services, and processes, as it transforms everyday life. Recent studies have increasingly examined how consumers respond to various forms of AI robots. This study adds to the existing literature by investigating the factors that drive consumers to adopt AI robots and the underlying psychological processes. Through three studies, we show that consumers' perceived congruence in an AI robot plays an important role in determining their adoption intention for the AI robot. We further demonstrate that such a perception of congruence is jointly determined by the AI robot's benefit type and appearance familiarity. The findings suggest that consumers are more inclined to adopt an AI robot when they are designed to provide hedonic benefits and resembles real-life objects. By contrast, an AI robot designed for utilitarian benefits are more likely to be adopted when it has the appearance of a machine. In addition, we show that perception of fit mediates this joint effect. Furthermore, we show that the effect of congruence perception on adoption intention is attenuated when consumers have a higher level of dialectical thinking. This study discusses both the theoretical and practical implications.