AI Assistant in Online Pharmacy

Abstract

Artificial intelligence (AI) has been increasingly popular in diagnosing diseases and recommending drugs in digital healthcare platforms. Leveraging the introduction of an AI-powered medical assistant to one drug category in an online pharmacy platform, we investigate how the adoption of AI affects users’ purchase behaviors using a difference-in-differences design. We find that the adoption of the AI assistant significantly increases users’ purchases in the platform, even for drugs not recommended by the AI assistant. Furthermore, we find that the positive effect of the AI assistant adoption is stronger for early technology adopters, inexperienced users, and users with higher privacy concerns, likely because these users tend to perceive higher value from AI. Finally, our mediation analysis shows that the AI feature increases users’ purchases by increasing their engagement levels in the platform. Our results have important implications for designing and evaluating AI features in online platforms.