Sustaining tourists’ adoption of generative AI for travel planning: the role of information quality and AI literacy

Purpose This study aims to examine how perceived tourism information quality and artificial intelligence (AI) literacy shape tourists’ continuance intention to use generative AI for travel planning through perceived usefulness and satisfaction and investigates whether these relationships vary across technical proficiency, communication proficiency and critical evaluation. Design/methodology/approach Survey data were collected from a paid South Korean online panel via Macromill Embrain. Using non-probability purposive sampling with eligibility screening, 391 tourists who self-reported having used ChatGPT to obtain travel-related information for travel planning within the past six months provided valid responses. Data were analysed using SPSS 25.0 and the PROCESS macro. Findings Perceived tourism information quality – encompassing accuracy, sufficiency, reliability, usefulness, personalisation and timeliness – strengthens perceived usefulness, which increases satisfaction and subsequently fosters continuance intention through a serial mediation process. AI literacy functions as a capability-based boundary condition: under lower information quality, higher literacy helps users sustain perceived usefulness by critically appraising and iteratively refining outputs, whereas lower literacy heightens vulnerability to poor information and weakens continuance intention. Originality/value Sustained generative AI use in tourism depends not only on information quality but also on users’ capability to engage with, evaluate and iteratively refine AI-generated content. By integrating information quality with multidimensional AI literacy, this study clarifies post-adoption belief formation under epistemically uncertain generative-AI outputs and identifies capability-based boundary conditions for sustained use.