Tourism and Hospitality and Artificial Intelligence: Use Concentrated in Marketing, Planning Tools That Are Not Yet Reliable, and Unsettled Rules on Pricing Algorithms

Purpose. Tourism and hospitality are labour-intensive, seasonal and built from many small firms, and they depend on trust, language and personal service. This article asks where artificial intelligence stands in the industry in 2026 and where it may be by about 2035. Design and methods. A review combining the share of the tourism research literature that involves AI (OpenAlex, 2010–2025), Eurostat's survey of AI use in EU accommodation enterprises (2021–2025), a benchmark of language agents on realistic travel planning, a press-documented case of a robot-staffed hotel, a US appeals-court decision on hotel pricing software, and a study of machine translation on a trading platform. Claims taken from the literature were checked against the sources cited. The scenarios are judgements built on this evidence. Findings. The share of tourism works that mention AI rose from 0.8% in 2018 to 5.5% in 2025, 6.7 times as much, and language-model works from 26 to 639. In the EU, 21.3% of accommodation enterprises used AI in 2025, more than four times the 5.0% of 2021 but only just above the 19.9% for all enterprises, and 78.3% used none. Use is concentrated in marketing and sales (12.5% of enterprises) and text mining (13.1%), while only 3.0% used AI in production processes, and every barrier named in the survey was cited more often than across all sectors. Planning is the application that consumers meet first, and it is not yet reliable: in a 2024 benchmark of 1,225 realistic travel-planning requests, the best language agent produced a plan meeting all constraints in 0.6% of cases. Guest-facing automation has failed in practice, as when a Japanese robot-staffed hotel was reported to have retired about half of its 243 robots. For pricing, a US appeals court in 2025 affirmed the dismissal of an antitrust suit over hotels' use of shared revenue-management software, leaving the law unsettled. We expect AI to be routine in hotel marketing, communication and revenue management by 2028, reliable booking agents to depend on progress in constraint-following in 2029–2032, and the balance of power between hotels and platforms to be the open question. Originality. The article separates the marketing uses where firms have adopted AI from the operational uses where evidence is thin, uses a benchmark to show how far trip-planning agents are from reliable, and notes the unsettled legal status of algorithmic pricing.