Consumers' responses to personalized service from medical artificial intelligence and human doctors

Artificial intelligence (AI) is transforming healthcare, yet little is known about how consumers experience medical AI and make decisions about using these providers for follow-up care. We take an interdisciplinary approach that combines behavioral and neuroimaging experiments to determine how anthropomorphism and personalization influence consumers' well-being and their decisions about a provider for follow-up care. Study 1 finds that, regardless of the type of medical provider used, consumers are likely to experience a heightened sense of well-being after a highly personalized medical consultation. However, regardless of whether the first interaction is highly personalized or not, they prefer follow-up visits with a human doctor, as mediated through empathy. Study 2 finds that, regardless of the type of medical provider used, the anterior cingulate cortex shows greater activation in a highly (vs. low) personalized interaction, the activation of which is also correlated with consumers' willingness to revisit that provider. Furthermore, the temporal occipital fusiform cortex shows greater activation in consultations with a human doctor (vs. medical AI), regardless of the interaction's level of personalization. The level of activation is also correlated with consumers' willingness to revisit with the initial provider. The findings suggest that medical AI cannot currently replace human doctors and that human doctors remain consumers' choice for medical consultations and treatment.