When powerful artificial intelligence backfires
通过定性和定量研究，发现将人工智能描述为“强大”会增加消费者的隐私担忧，但这种效应在消费者与人类代理互动时减弱，在与AI互动时增强；提供隐私保障可缓解担忧并提高数据披露意愿。
Despite recent advancements in artificial intelligence (AI) in hospitality, little is known about its unintended consequences on consumers’ privacy concerns. Through an empirical package combining qualitative and quantitative evidence, this research reveals that framing AI as “powerful” enhances privacy concerns (Studies 1–5) by reducing AI-control over data (Study 3). Notably, such effects are reduced in consumer-human agent interactions and increased in consumer-AI ones (Study 4). Finally, interventions providing privacy guarantees can reduce privacy concerns stemming from powerful artificial agents and enhance willingness to disclose data (Study 5). Our findings highlight the unique issues arising from human-AI interactions when using powerful AI and their influence on consumers’ privacy concerns and provide practical implications for hospitality managers.