Artificial Intelligence Adoption in The Hospitality Industry: A Narrative Analysis of AI Deployment In Leading Global Hotel Groups

Abstract

Artificial Intelligence (AI) is increasingly important in hospitality. However, research on how major hotel chains document and manage their AI strategies using multi-source triangulation is scarce. Most previous studies rely exclusively on corporate data from a single source, thereby diminishing the quality of the evidence. This study fills that gap through a two-phase systematic narrative synthesis rooted in an interpretivist, inductive approach. In the first phase, a PRISMA-P-guided search across four academic databases and grey literature narrowed 402 records to 44 sources via iterative coding (Cohen’s Kappa = 0.81). The second phase involved a three-source triangulated micro-case analysis of four hotel groups (Hilton, Marriott, IHG, Wyndham), combining corporate communications, independent trade press, and academic/practitioner sources. Findings are organised using the People-Process-Technology (PPT) model alongside TAM, UTAUT, and DOI theories, serving an interpretive rather than a statistical purpose. AI adoption was mapped across six operational domains, revealing four tentative patterns: accessibility-led, hybrid human-AI, broad automation, and governance-led. Eight recurring challenges were identified, such as workforce displacement, algorithmic bias, and regulatory fragmentation, and categorised by PPT components. The main contribution is both methodological and synthetic: it uses structured triangulation, here for the first time in this field, to reduce promotional bias from single-source data. It also provides an initial, evidence-based vocabulary and practitioner recommendations, although it does not propose a validated theory. Since the evidence is limited to four publicly listed, purposively chosen hotel groups and publicly available sources, the archetypes and framework are presented as hypotheses for future research. This limitation is explicitly addressed throughout the paper, not just in the conclusion.