How AI shapes the cultural tourism visitor experience: A scoping review with a personalisation lens.

Abstract

This scoping review maps how artificial intelligence, particularly large language models (LLMs) and adjacent modalities, influence the cultural tourism visitor experience through personalisation. Following Arksey and O’Malley’s PRISMA-ScR reporting guidelines, searches were conducted in Scopus and Google Scholar (2015–2024), yielding 610 records, of which 18 studies were retained after screening. The analysis identifies a layered “personalisation stack” comprising conversational guidance (LLM pilots), knowledge-graph storytelling, behaviour-aware recommenders, telemetry-informed orchestration, and immersive/assistive media. Reported benefits cluster around engagement, perceived relevance, comfort in VR, and progress towards accessibility. However, value is contingent on governance, autonomy calibration, spatial/crowding effects, and explainability. Evidence gaps include longitudinal evaluation, multimodal LLM integration with knowledge graphs and sensor streams, accuracy/bias auditing, privacy-by-design, and performance reporting for real-time operation. The paper synthesises theoretical, managerial, and policy implications and outlines a practice-oriented checklist. Overall, the findings clarify how AI shapes cultural visitor experiences and the safeguards required to deliver equitable value at scale.