Human or AI? Identity Salience Enhances Historically Marginalized (but Not Privileged) Consumers’ Choice of Algorithmic Service Providers

When their identity is salient, consumers from historically marginalized and privileged groups are differentially influenced when it comes to choosing algorithmic service providers (e.g., AI-enabled tellers, chatbots, robots, and digital kiosks). Under identity salience, as would be the case in a service context where the service providers are all from historically privileged groups, consumers from historically marginalized groups (e.g., Black people) become concerned about identity-based social judgment, leading to relatively higher preference for algorithmic service providers. Underrepresentation of historically marginalized groups in white-collar professions (e.g., financial services), regulatory pressures to abandon diversity hiring initiatives, and a tight labor market make it increasingly difficult for companies to ensure diversity among frontline employees. How can managers provide consumers from historically marginalized groups the services they seek in nondiverse settings while balancing other constraints and reducing identity-based social judgment concerns? The results of a series of field studies and controlled experiments suggest that algorithmic service providers can provide a means to serve historically marginalized consumers in nondiverse service settings. The effect attenuates in service contexts where historically marginalized consumers are less concerned about identity-based social judgment. Interestingly, identity salience does not influence historically privileged consumers’ choice of algorithmic service providers.