FACULTÉ & RECHERCHE

Although artificial intelligence (AI) outperforms humans in many tasks, research suggests some consumers are still averse to having AI perform tasks on their behalf. Informed by the literature of customer decision-making process, we propose and show that consumer autonomy is a significant predictor of customers' decision to adopt AI in the purchasing context. Across three experiments, we found that the delegation of purchasing tasks to AI, which restricts choice and decision dimensions of consumers' perceived autonomy, reduces the likelihood of AI adoption. Our results show that the effects of choice and decision autonomy on AI adoption holds even when product choice evaluation is complex. We also found that identity-relevant consumption moderates this relationship, such that it interacts with choice and decision autonomy. Specifically, despite lacking choice and decision autonomy, those who identify strongly with a given activity are more likely to use an AI-enabled app to purchase the product needed to perform this activity. With these insights into when and why consumers are likely to use AI-enabled technology, firms might effectively increase its adoption.