Rethinking human–AI agency in tourism and hospitality: A framework for the agentic AI era

This study critically examines the ontological assumptions underlying artificial intelligence (AI) research in tourism and hospitality, with growing attention to agentic AI. Existing literature predominantly conceptualizes AI as a tool, limiting discourse on how tourism experiences and sociotechnical systems are co-constructed through human–AI interactions. To address this gap, this study adopts a multi-method approach, combining bibliographic coupling, large language model (LLM)-assisted extraction of theories and variables, and expert interviews. Drawing on digital humanism, transhumanism, and posthumanism, it investigates how assumptions about human–AI relationships are embedded in existing studies. The findings reveal that tourism AI research remains grounded in instrumental human-centeredness framework, conceptualizing AI as an instrument for assisting or augmenting human capabilities. However, industry practice reveals AI as hybrid assemblages entangled with human labor, platforms, and data infrastructure. This study calls for a more relational, multi-layered understanding of human–AI interactions and contributes a framework comprising 15 theoretical propositions.