Unseen, Unbothered? - The Hidden Impact of AI Disclosure on Consumers’ Intention to Adopt Recommendations

Sammanfattning

This study explores the behavioral mechanism of consumers’ intention to adopt AI-generated product recommendations in the e-commerce setup with the condition of labeling the usage of AI in the system. Using an experimental design with 165 university students, the results show that explicit labeling recommendations as AI-generated significantly reduces the intention to adopt the recommendations. This effect is fully mediated by psychological reactance, indicating that labeling the recommendations as AI-generated triggers a perceived loss of autonomy, leading to resisting to adopt. The study also finds that having a positive attitude towards AI reduces the negative effect of labeling, on the other hand, perceived accuracy of recommendations does not have any significant impact on the negative effect that labeling has on the intention to adopt. These findings offer important implications for marketers and policymakers aiming to balance transparency and user experience in an AI-driven environment.