Artificial intelligence (AI) is a cost-efficient innovation that challenges customers' consumption patterns and fears of uncertainty. This study assesses whether the likelihood that consumers adopt AI in banking services depends on tastes across different cultures. We propose a culturally-augmented Arrow-Bilir-Sorensen model to assess the propensity that consumers use AI. Analyses of a unique ING Bank dataset encompassing 11,000 respondents from 11 countries reveal that success rates for the diffusion of robo-advisory financial services in retail banking vary substantially due to the cultural boundedness of choice. This bias seems to be associated with social capital rather than the fear of novelty.