Generative AI drives rapid expansion of tourism and hospitality research

Cluster analysis of keywords in generative AI papers reveals four distinct research streams: ChatGPT’s role in tourism education and trust-building; the ethical implications and adoption patterns of generative AI; its influence on travel behaviour and perceived risk; and technical focus areas such as chatbots and natural language processing. ChatGPT emerged as the most frequent keyword, appearing in 41 of the 50 generative AI documents, followed by artificial intelligence, tourism, and generative AI as recurring terms.

A newly published review confirms a significant shift in artificial intelligence research within the tourism and hospitality sector, marking the rise of generative AI as a central research theme. The peer-reviewed study, titled "Artificial Intelligence Research in Tourism and Hospitality Journals: Trends, Emerging Themes, and the Rise of Generative AI," was published this week in the journal Tourism and Hospitality. It systematically maps AI research trends by analyzing 921 academic publications indexed in Scopus, with a specific focus on machine learning, deep learning, sentiment analysis, and most recently, generative AI tools like ChatGPT.

Through performance analysis and science mapping using tools like VOSviewer and Biblioshiny, the research reveals both quantitative growth and qualitative shifts in academic focus.

How AI-related research in tourism and hospitality has evolved over time

Publication trends illustrate an explosive growth in AI-related research from 2018 onward, increasing from fewer than 24 papers annually before that year to a peak of 277 publications in 2024. This trajectory underscores AI's increasing integration into both theoretical and practical aspects of tourism and hospitality. Early research efforts emphasized neural network models and expert systems to forecast tourism demand, while more recent work pivots toward service automation, social media analytics, and large language models.

The review identifies R. Law of the University of Macau as the most prolific contributor, with 38 publications and 2,498 citations. China leads as the top-producing country, contributing 262 of the 921 documents, followed by the United States and the United Kingdom. The Hong Kong Polytechnic University ranked as the most productive institution. Tourism Management, The International Journal of Hospitality Management, and The International Journal of Contemporary Hospitality Management are the leading journals publishing AI-focused articles.

Keyword co-occurrence analysis uncovered four dominant clusters: machine learning and sentiment analysis of online reviews; adoption of AI - including ChatGPT and robotics - in hospitality; artificial neural networks for tourism management; and random forest models applied in travel demand forecasting. These thematic clusters demonstrate the diversification of AI techniques being deployed to understand and enhance consumer experience, improve service design, and optimize operational decision-making.

Role of generative AI

A critical segment of the study focuses on the emergence of generative AI. Of the 921 total publications, 50 - comprising 48 articles and 2 reviews - directly address tools such as ChatGPT, with a sharp increase in volume since 2023. Notably, Carvalho and Ivanov’s 2024 paper “ChatGPT for Tourism: Applications, Benefits and Risks” has already garnered 208 citations, making it the most influential publication in this new subfield. Other heavily cited papers explore ChatGPT’s use in education, automation strategies, and personalized service development, reflecting growing interest in both opportunities and challenges presented by these technologies.

Cluster analysis of keywords in generative AI papers reveals four distinct research streams: ChatGPT’s role in tourism education and trust-building; the ethical implications and adoption patterns of generative AI; its influence on travel behaviour and perceived risk; and technical focus areas such as chatbots and natural language processing. ChatGPT emerged as the most frequent keyword, appearing in 41 of the 50 generative AI documents, followed by artificial intelligence, tourism, and generative AI as recurring terms.

The shift from early AI models like neural networks to sophisticated language models aligns with a broader industry trend toward intelligent automation. The COVID-19 pandemic catalyzed much of this innovation, prompting the integration of AI-powered service robots and chatbots to reduce human contact and ensure business continuity. This technological acceleration is echoed in highly cited works examining robotics adoption in hotels, consumer perceptions of chatbot intelligence, and AI’s role in improving hygiene and service efficiency.

Implications of generative AI

Beyond mapping trends, the study highlights the implications of generative AI for academic and industry stakeholders. AI is now regarded as a strategic asset for tourism enterprises, enabling data-driven decisions in pricing, resource allocation, and marketing. It also presents transformative potential in shaping traveller expectations, destination competitiveness, and employee reskilling needs. Researchers argue that tourism businesses adopting generative AI tools can deliver more personalized, efficient, and scalable service models.

However, the study emphasizes the importance of transparency and ethical deployment of these technologies. While automation can boost productivity, it must be balanced with customer trust, data privacy, and accountability. The authors call for responsible AI adoption practices and greater interdisciplinary collaboration to explore the long-term effects of generative AI in tourism contexts.

Methodologically, the research leverages bibliometric techniques to map large-scale data and reveal evolving scholarly patterns. Unlike systematic reviews limited to small document sets, this study utilizes science mapping to process nearly a thousand publications, providing a high-level view of thematic development across decades. Researchers advocate for replicating this approach in related disciplines such as business administration and marketing, where AI is also reshaping research agendas.

Lastly, the study also acknowledges certain limitations. For instance, reliance on Scopus metadata may introduce omissions or inconsistencies due to keyword variation. Additionally, the rapidly evolving nature of AI technologies means that bibliometric snapshots can quickly become outdated. The authors suggest ongoing updates and parallel analyses using other databases such as Web of Science or Dimensions for cross-validation.