Artificial intelligence adoption in hospitality: challenges and opportunities across hierarchical levels Open Access

In recent years, technological advancements have become integral to human life, and, notably, the integration of artificial intelligence (AI) has emerged as a prevalent trend in this technological era (Özen and Özgül Katlav, 2023; Xu et al., 2024). AI technology has permeated diverse sectors, including engineering, banking, medical treatment and hospitality, where it collaborates with or supplants human roles (Wong et al., 2023). Particularly, AI adoption within the services sector, particularly in the labor-intensive hospitality industry, has witnessed rapid growth, marked by continuous advancements (Ghesh et al., 2024; Huang and Zheng, 2023).

The hospitality industry stands as the most rapidly expanding sector globally, yielding an estimated $8tn in revenue and facilitating the creation of 292 million jobs on a global scale (Ruel and Njoku, 2020). The advent of AI applications is reshaping prevailing business paradigms, prompting novel opportunities and challenges for the hospitality domain. Hence, in the contemporary landscape of the hospitality industry, the integration of AI technologies has emerged as a pivotal strategy to enhance operational efficiency, elevate guest experiences and drive competitive advantage. However, the adoption of AI within the hotel sector is not short of challenges, as it necessitates navigating multifaceted barriers to achieve user acceptance (Morosan and Dursun-Cengizci, 2024).

Scholarly attention has predominantly focused on evaluating the costs and benefits associated with AI implementation in the hospitality sector. Notably, AI streamlines tasks such as self-check-in/out, room services, housekeeping, concierge services and chatbot interactions, enhancing efficiency while reducing human costs (Ersoy and Ehtiyar, 2023; Zhu et al., 2023). A distinct research avenue examines costumer acceptance or rejection of AI technology, with findings indicating varying levels of convenience and efficiency perception among costumers due to technological complexity and lack of knowledge (Rasheed et al., 2023; Vorobeva et al., 2023). Although current studies primarily analyze the impacts of AI from the standpoint of customers, there is a justified need and calls for additional exploration into its effects on employees (Leung, 2019; Li et al., 2022; Rasheed et al., 2024). Thus, it is imperative to explore the implications of AI technologies on hospitality employees, shedding light on emerging opportunities and threats to guide competency development aligned to evolving technology.

This paper examines how artificial intelligence adoption is perceived across different hierarchical levels within the hotel industry, focusing on the specific challenges and opportunities involved. The study includes 55 participants representing a range of roles, from top managers and first-line managers to nonmanagerial employees. Using semi-structured interviews and a consensus mapping approach (Tarakci et al., 2014), the research aims to capture and compare the perceptions of these groups regarding the key barriers and drivers influencing artificial intelligence adoption. The conceptual framework guiding this study is presented in Figure 1.

The contribution of this study is threefold. First, categorizing participants into three groups facilitates a nuanced examination of diverse viewpoints on AI adoption within the hotel industry. Indeed, our study captures a comprehensive spectrum of perspectives, facilitating a holistic understanding of complexities surrounding AI integration within groups of workers occupying different hierarchical positions. Second, through rigorous analysis of interview data and consensus mapping techniques, this study seeks to distill key themes, patterns and divergences in perceptions between hierarchical levels, thereby identifying salient factors that inhibit or propel AI adoption within hotel operations. Third, recognizing the importance of effective communication in organizational change, our study proposes tailored communication strategies to increase awareness and foster acceptance of AI technologies among different hierarchical groups. All in all, our findings enrich the literature on AI adoption in the hotel industry by offering insights into diverse hierarchical perspectives and suggesting strategies to overcome adoption challenges and to seize emerging opportunities of this transformative process (Buhalis, 2025). Its empirical findings and recommendations aim to guide managerial decisions, fuel academic discussions and encourage AI-driven innovations in hospitality management.

The rapid advancement of technology has transformed the operational dynamics of the hotel industry, driving widespread automation and enhancing the capacity of hospitality businesses to deliver value to customers (Jabeen et al., 2022). Service automation within hotels encompasses a broad spectrum of applications, including automated check-in and check-out systems, artificial intelligence-powered chatbots for customer service, robotic concierge services and data-driven personalization of guest experiences.

To analyze how employees might perceive the adoption of artificial intelligence in the workplace, this study draws upon Elton Mayo’s Human Relations Theory (Sarachek, 1968) as a conceptual framework. While Mayo’s theory does not specifically address technological advancements such as artificial intelligence, its focus on human dynamics, interpersonal relationships and the importance of employee well-being provides a valuable lens through which to interpret responses to technology-driven changes. According to Mayo, employees’ sense of belonging and the quality of workplace relationships significantly influence job satisfaction and productivity. Applying this perspective to modern technological contexts, employees might fear that artificial intelligence could replace interpersonal interactions or erode their sense of connection within the organization.

Recognizing these concerns is crucial to ensuring that technological efficiency does not come at the expense of meaningful human connections in the workplace (Zhong et al., 2022). Creating an environment where artificial intelligence is perceived as a tool to enhance, rather than replace, human contributions can help alleviate fears and build trust. Clear communication about the supportive role of artificial intelligence can reinforce the message that technology complements, rather than threatens, employees’ professional roles. In addition, fostering positive peer influence and involving informal leaders in advocating for artificial intelligence adoption can reduce resistance and promote workforce alignment.

Mayo’s emphasis on the value of human relationships underscores the need to integrate empathy and collaboration into artificial intelligence implementation strategies. This approach is essential for maintaining employee motivation and engagement during transitions to automated systems. While extensive research has focused on consumer perceptions of service automation, significantly less attention has been given to managerial attitudes toward these technologies (Leung, 2019). This gap in the literature is particularly critical, as managers are key decision-makers who influence how and whether automation technologies are effectively implemented within organizations. Understanding managerial perspectives is therefore vital for identifying the barriers and drivers of technology adoption and for shaping the strategic considerations that guide these decisions.

This study aims to contribute to the expanding research on service automation by addressing the underexplored area of managerial attitudes. By examining how key hotel stakeholders – including owners, managers and employees – perceive the benefits and challenges associated with adopting automation technologies, this research offers a nuanced understanding of the factors that either facilitate or hinder technological innovation in the hospitality sector.

Following Rasheed et al. (2024), adoption drivers of AI in hospitality can be categorized into three distinct groups: operational, emotional and situational. Operational drivers refer to the tangible advantages derived from the fundamental capabilities of AI technologies. Scholarly investigations have consistently emphasized the pivotal roles of perceived usefulness in shaping consumers’ attitudes toward AI adoption (Park et al., 2021). For instance, AI facilitates humans in efficiently accomplishing various tasks, including self-check-in/out, housekeeping, concierge services and chatbot interactions for information retrieval (Law et al., 2023; Li et al., 2021; Wong et al., 2023; Zhu et al., 2023). Hence, operational efficiency pertains to the tangible benefits stemming from the core features of AI technologies, specifically in terms of optimizing processes. Furthermore, AI adoption correlates with reduced workloads and heightened productivity among employees (Buhalis and Moldavska, 2022; Ersoy and Ehtiyar, 2023).

Tasks such as data analysis, standardized customer service and administrative duties are streamlined, allowing staff to focus on higher-value activities. Indeed, prior studies underscore AI technologies’ ability to optimize inventory management and streamline processes, leading to cost savings for businesses. In addition, factors such as perceived interactivity and innovativeness of AI technologies positively contribute to intentions to adopt and repurchase these technologies (Go et al., 2020; Pillai and Sivathanu, 2020). More specifically, AI technologies’ ability to deliver personalized experiences based on individual preferences and behaviors fosters positive attitudes toward AI adoption, as users perceive the technology as valuable and relevant to their customization needs. Indeed, service customization triggers greater customers’ knowledge, emotional attachment and behavioral commitment (Buehring and O’Mahony, 2019).

Shifting the focus away from operational benefits for businesses, emotional drivers explore the complex domain of consumer sentiments and motivations toward AI adoption. Prior studies underscore the relevance of perceived human likeness and intrinsic motivations, moderated by sociodemographic aspects such as age, gender and income in shaping consumers’ propensity toward AI adoption in the hospitality sector (Belanche et al., 2020). AI technologies streamline processes (Rasheed et al., 2024), provide personalized experiences (Go et al., 2020) and offer real-time support (Li et al., 2021), enhancing overall service quality. Hence, studies suggest a positive correlation between AI adoption and heightened customer satisfaction levels, underscoring the importance of AI-driven innovations in meeting evolving consumer demand (Alam et al., 2023).

Focusing on the macro level, situational drivers are facilitating conditions that are key contextual drivers in AI adoption intentions. The macro context where users interact with AI technologies has a considerable impact on their perceptions toward AI value and usefulness, ultimately either pushing or inhibiting adoption behaviors (Lin et al., 2020; Mariani and Borghi, 2023). Both mimetic and first-mover pressure relate to the contextual factors influencing user attitudes and behaviors toward AI adoption, specifically in terms of gaining a competitive edge through the use of AI technologies.

Barriers toward AI adoption in hospitality are mainly grouped into value, risk and usage barriers (Rasheed et al, 2024). When users perceive the cost of adopting a new technology outweighing the potential benefits, they may view it as an offering of less value compared to existing alternatives. Hence, value barriers may arise, creating reluctance to adopt the new technology or hindering its acceptance (Laukkanen et al., 2008). High implementation costs (i.e. substantial upfront costs for technology acquisition, integration, training) and concerns regarding return on investment are considered value barriers due to their potential to negatively influence user perceptions of a new product or service (Rasheed et al., 2023).

Prior studies unveil the emergence of risk barriers, referring to the degree of uncertainty and associated risks inherent in novel products, processes, services or technologies, shaped by user perceptions or experiences (Chen and Kuo, 2017). Privacy and security concerns are categorized under risk barriers, revolving around uncertainties about regulatory compliance like General Data Protection Regulation (GDPR) to uphold customer trust (Rasheed et al., 2023).

While few studies elucidate no significant relationship between technological anxiety and the intention to adopt AI (Pillai and Sivathanu, 2020), other scholars tend to acknowledge usage barriers that include a mismatch between the new technology and the users’ prior experiences (Antioco and Kleijnen, 2010). For instance, existing research highlights factors such as the lack of awareness and understanding that reflect the incompatibility of AI technology with user’s existing experiences, habits and acceptance standards. Businesses may not fully comprehend the potential of AI and its applications, indicating a lack of awareness as a barrier to adoption (Alam et al., 2023; Huang et al., 2022). Furthermore, Li et al. (2019) show a significant positive correlation between awareness of AI and robotics and employees’ intention to turn over. This underscores the reluctance to embrace AI technologies due to fear of job loss, changes in work processes or perceived threats to professional roles. Consequently, resistance to change signifies the incongruity between AI and current processes and roles, impeding adoption. In addition, Rasheed et al. (2023) classify technological complexity as a usage barrier, encompassing challenges related to AI service complexity and the requisite technical expertise often being absent within organizations. Technical complexity represents a usage barrier to AI adoption due to difficulties in implementing and integrating AI solutions into existing systems and workflows.

To explore the challenges and opportunities associated with artificial intelligence adoption in the hotel industry, this study used semi-structured face-to-face interviews with 55 key information-rich stakeholders occupying diverse hierarchical positions within hotels. The primary focus was on hotels located in the South Tyrol region, selected due to the established collaboration with local hotel associations, which facilitated access to a relevant and engaged sample. The South Tyrol area is characterized by a high concentration of family-run, small to medium-sized hotels primarily targeting leisure clients. This homogeneity in hotel type and market focus contributed to a more consistent and comparable data set, enhancing the internal validity of the study. In addition, we strategically focused the study on South Tyrol due to its strong alignment with the study’s objectives and its representativeness of the broader leisure-oriented, family-owned hospitality segment, which embodies many of the distinctive characteristics of Italian hospitality. As a result, this regional focus serves as a microcosm of Italian hospitality, offering a representative model that allows the study’s findings to provide valuable insights into the national context.

Following the managerial hierarchy framework established by Robbins et al. (2020), participants were classified into three categories based on their roles:

1. top managers;

2. first-line managers; and

3. nonmanagerial employees.

Individuals such as hotel general managers, hotel owners and CEOs were categorized as top managers. Those occupying roles including heads of the booking department, maîtres d’hôtel, chefs, heads of housekeeping, heads of marketing and communication, heads of procurement and inventory and chief maintenance officers were classified as first-line managers. Finally, receptionists, bartenders, servers, dishwashers, kitchen assistants, maintenance staff and housekeeping staff were identified as nonmanagerial employees. The interviews involved n. 27 top managers, n. 13 first-line managers, n. 15 nonmanagerial employees. They lasted on average 24 min and were conducted between September 2023 and March 2024.

The questionnaire consists of three sections. Section 1 includes general open-ended questions such as “What kind of experience do you have with AI in the hospitality context?”, “Where/How do you envision applications of AI in the hotel sector in the coming years?” and “What are some empirical examples or implementations of AI in the hospitality context?”. Section 2 delves into more specific topics, addressing barriers and drivers of AI adoption in the hotel industry. Sample questions include: “What are the main barriers and drivers toward AI adoption in the hospitality context?” and “Do you believe the benefits outweigh the obstacles, or vice versa?”. Finally, participants were asked to prioritize five groups of barriers and five groups of drivers related to AI adoption identified from the literature. Data from the first two sections were analyzed using an open coding procedure (Miles and Huberman, 1994), which involved identifying similarities and differences between quotations and grouping similar responses into categories. The emergent themes were then compared with categories identified in the literature to support the existing literature-based categories.

From the literature review, five main themes relative to drivers emerge, such as (i) operational (i.e. operational efficiency [OE] e.g. Buhalis and Moldavska, 2022; service personalization [P] e.g. Buehring and O’Mahony, 2019; cost saving [CS] e.g. Go et al., 2020), (ii) emotional (i.e. enhanced customer satisfaction [SAT] e.g. Alam et al., 2023) and (iii) situational (i.e. competitive advantage [CA] e.g. Mariani and Borghi, 2023). Likewise, from the literature, also different groups for barriers emerge, such as (i) value barriers (i.e. cost concerns [CC] e.g. Rasheed et al., 2023), (ii) risk barriers (privacy and security concerns [PS] e.g. Rasheed et al., 2023) and (iii) usage barriers (i.e. resistance to change [RC] e.g. Li et al., 2019; technological complexity [TC] e.g. Rasheed et al., 2023; lack of awareness and understanding [LAU] e.g. Huang et al., 2022) emerge. Study participants were asked to prioritize the five categories related to the drivers and barriers by ranking them in order of importance, resulting in a personal ranking of factors that propel/inhibit AI adoption that ranged from the most important driver/barrier to the least important driver/barrier. Responses were used to assess the degree of consensus within and between diverse hierarchical groups (i.e. top manager, first-line manager, nonmanagerial employees) via Tarakci et al. (2014) consensus mapping. Consensus mapping by Tarakci et al. (2014) is particularly effective with limited sample sizes due to its reliance on the aggregation of expert opinions and qualitative insights (depth) rather than statistical generalization (breadth). The method emphasizes structured dialogue and collaborative evaluation, allowing researchers to distill shared priorities and key themes even from small, diverse groups. Its focus on capturing depth and consensus ensures meaningful results, bypassing the constraints of large-sample quantitative techniques.

Descriptive statistics of the sample are presented in Table 1.

Top managers, first-line managers and nonmanagerial employees assigned priorities to five different categories of drivers and barriers. We begin the analysis by exploring the preference orderings of the different respondent groups. To achieve this, we use heatmaps to visualize the correlations among the main categories of drivers and barriers identified in the interviews. The correlation analysis indicates a lack of significant relationships between the response categories for both managerial and nonmanagerial respondents. In other words, neither group perceives meaningful intrinsic connections between the variables, instead viewing them as independent phenomena influenced by the adoption of artificial intelligence tools. We therefore require a different set of tools to examine the consistency of the evaluations provided by the different groups of respondents when ranking the different categories of drivers and barriers.

Two measures of consensus have been computed to analyze the responses received (Cozzio and Furlan, 2023). The first one focuses on the degree of consensus within the members of each group and follows from a dimension reduction technique based on principal component analysis (PCA). The second measure analyzes consensus between groups and is based on classic multidimensional scaling. Both measures were introduced in the literature by Tarakci et al. (2014). All computations have been performed with MATLAB.

The degree of consensus within each group was obtained by applying the vector model of unfolding (VMU) as defined by Tarakci et al. (2014), which transposes the standard data matrix used in PCA (Borg and Groenen, 2005). Intuitively, PCA is a dimension reduction technique that computes the main components explaining the variance of a series of characteristics – composing the columns of a matrix – that are used to describe a set of alternatives – place in the corresponding rows. These components can be represented as the axes of a Cartesian plane where the importance of the characteristics describing the alternatives is defined by their relative position in the plane. Each alternative is then represented in the plane to illustrate the relative importance of each component in their evaluation. Tarakci et al. (2014) apply PCA to a matrix where respondents are shifted to define the columns while the categories ranked are placed in the rows.

The formal environment defined by these authors has been consistently implemented to evaluate the strategic interactions taking place among the members of different groups within organizations and the effect that consensus has on their subsequent performance (Wei et al., 2025; Ateş et al., 2020). Willems and Meyfroodt (2024) provide a recent review of the literature dealing with consensus analysis and the strategic interactions taking place among the members of the groups.

From a technical viewpoint, consider the standardized data matrix H that lists the m categories evaluated in its rows and the respondents along its n columns. The VMU method proposed by Tarakci et al. (2014) in p dimensions is based on the minimization of the sum of the squared errors derived from H and the low-dimensional representation XA′ and defined by $LVMU(X,A)=‖H−XA′‖2=∑ijeij2$ where X is an m × p matrix of object scores and A is an n × p matrix of component loadings for the first p components.

The component loadings in A are the correlations between the object scores for each category and the evaluations of each respondent. Tarakci et al. (2014) define the consensus within each group as the length of the average component loading vectors in A across its first two components $α=∑p=12(∑iaipn)2$⁠.

where a_ip denotes the p-th component loading for respondent i. As is also the case in PCA, the results obtained can be represented in the Cartesian plane using a biplot. The categories are described in order of preference relative to the horizontal axis, while the respondents composing each group are depicted as vectors.

The correlation between respondents can be approximated via the cosine of the angle between their vectorial representations (Linting et al., 2007). Larger angles describe less similar evaluations, while the opposite is true for smaller angles. The same intuition applies to the consensus within the group, with tighter clusters of vectors demonstrating a higher degree of consensus.

The preferences of the different respondents are represented by the orthogonal projection of each category item on their corresponding vectors. The farther an item is projected into the vector, the more preferred it is, while the items projected in the opposite direction are less preferred. In this regard, the horizontal axis corresponds to the prototypical member representing the opinion of the group. That is, the projection on the horizontal axis represents the overall ranking of the group.

Figures 2 and 3 illustrate the biplots defining the degree of within-group consensus for each group of respondents when considering drivers and barriers, respectively.

In particular, the projections of the categories on the horizontal axes of Figure 2 describe the importance assigned by top managers (operational efficiency, competitive advantage and satisfaction), first-line managers (satisfaction, personalization and operational efficiency) and nonmanagerial employees (cost saving, operational efficiency and satisfaction) to the corresponding drivers. Clearly, the preferences of each group differ considerably regarding the relative importance of the categories evaluated.

A similar intuition is derived from the barriers, though some basic degree of consensus can be observed among top and first-line managers, who consider lack of awareness and understanding and resistance to change as the main two barriers – though they differ in their relative importance between both groups of respondents. Interestingly, privacy and security remains a minor concern for top and first-line managers, and even nonmanagerial employees rank it quite low when considering the barriers impeding the adoption of artificial intelligence in the hotel industry.

It must also be noted that preferences are not substantially consistent among the members of each group. As illustrated in Tables 1 and 2, within-group consensus is quite low for all groups, remaining below 51% when considering both drivers and barriers – with top managers displaying the lowest consensus in both cases. The subsequent ranking differences across groups will be examined in more detail through the next section (Table 3).

Consensus between groups is measured in terms of the correlation exhibited by the ranking preferences of the prototypical group members who represent better the whole group’s opinion. More precisely, the measure proposed by Tarakci et al. (2014), denoted by r (A, B), is determined by the correlation of the object scores of the categories on the first principal component for respondent groups A and B. The higher this value, bounded between zero and one, the higher the consensus between both groups. The results derived from the pairwise comparisons performed between groups across scenarios are presented in Table 4.

The distance between the groups of respondents defining the symmetric matrix of correlations is computed using classical multidimensional scaling (MDS). The corresponding output obtained is represented in Figure 4:

Tarakci et al. (2014) suggested defining ten rings to describe the difference in correlations relative to the group placed at the center of the plot. In this regard, a higher level of consensus is observed as the distance between points decreases, representing a more aligned evaluation between groups. In our graphical representation, top managers are at the center of the MDS plots. The distance between the bubbles shows the degree of consensus between the groups. The sizes of the bubbles denote the degree of the within-group consensus in each group (α), and the rings that surround the bubbles depict the size of a bubble when there is perfect consensus within a group (α = 1).

Between-group consensus differs markedly across groups of respondents. First-line managers and nonmanagerial employees display substantial differences with respect to top managers when considering the drivers. Consensus between groups increases when considering the barriers, though the correlations between groups remain quite low also in this case.

This study provides a pioneering perspective on artificial intelligence (AI) adoption within the hospitality industry by shifting the focus from a predominantly customer-centric view (Wong et al., 2023; Huang and Zheng, 2023) to an employee-centered approach. While previous studies highlighted how guests often perceive automated services as enhancing convenience and efficiency, concerns over reduced human interaction and data privacy continue to present significant challenges (Zhong et al., 2022). This study reinforces the importance of maintaining a balance between technological efficiency and the human touch that defines hospitality, aligning with recent calls to address the evolving role of AI in hospitality and tourism marketing (Bulchand-Gidumal et al., 2024).

Second, this research contributes to the theoretical discourse by examining managerial and employee perspectives on AI adoption across three distinct hierarchical levels: top managers, first-line managers and nonmanagerial employees. The nuanced approach reveals that while operational drivers such as operational efficiency, cost-saving and service personalization are universally valued (Ersoy and Ehtiyar, 2023; Zhu et al., 2023), distinct priorities emerge within each group. Top managers focus on competitive advantage, aligning with strategic goals, whereas nonmanagerial employees prioritize cost savings due to their operational roles. First-line managers, positioned between these extremes, emphasize personalized service, demonstrating an integrative perspective that balances internal efficiency with external customer satisfaction (Figure 5).

Building on recent literature, including the evolution of artificial empathy in the hospitality metaverse era (Assiouras et al., 2025), this study highlights the importance of tailoring artificial intelligence strategies to different hierarchical roles within organizations. For instance, nonmanagerial employees, who are deeply involved in daily operations, have a unique perspective on how artificial intelligence tools can streamline processes through cost saving. For example, receptionists using artificial intelligence-driven chatbots can experience firsthand the benefits of faster customer interactions and smooth workflows. Our study’s findings (Figure 5) unveil an organizational continuum where cost saving (i.e. inward orientation), favored by nonmanagerial employees, and competitive advantage (i.e. outward orientation), prioritized by top managers, stand as opposing ends of the spectrum. Positioned in the middle of these extremes, the possibility to offer personalized experience emerges as a pivotal driver for first-line managers, integrating both internal (i.e. artificial intelligence impact on operations) and external (i.e. artificial intelligence impact on customers) orientations. In addition, this study complements the existing knowledge about barriers that prevent the acceptance of artificial intelligence (Rasheed et al., 2023; Vorobeva et al., 2023). Intraorganizational inhibitors such as resistance to change and lack of awareness and understanding are deemed most critical across hierarchical levels, necessitating collaborative efforts to alleviate fears of job displacement and enhance artificial intelligencés perceived value. Conversely, interorganizational inhibitors such as privacy and security concerns hold marginal significance in impeding technological acceptance.

Third, this study builds on emerging research on anthropomorphism-based artificial intelligence in hospitality (Saputra et al., 2024), highlighting that artificial intelligence should not only optimize operations but also enhance employees’ sense of belonging and acceptance. By integrating human-like elements into artificial intelligence tools, organizations can create a more supportive work environment. Importantly, the study offers valuable insights into identifying and leveraging human-preferred elements of artificial intelligence, promoting a balanced blend of technological innovation and human-centric hospitality values. Thus, our study underscores the potential of anthropomorphic artificial intelligence to facilitate smoother adoption, enhance workplace cohesion and strengthen the integration of the technology within human-driven service environments.

Fourth, this study enhances the understanding of barriers to technological acceptance by applying Elton Mayo’s Human Relations Theory (Sarachek, 1968). While Mayo did not address technology directly, his focus on human interactions and employee well-being offers a useful framework for managing organizational change. The study highlights the importance of effective, tailored communication strategies (Morosan and Dursun-Cengizci, 2024) to build trust, increase awareness and promote technology acceptance across different hierarchical groups.

This study provides valuable managerial insights into effectively implementing artificial intelligence (AI) within the hospitality industry by considering the distinct needs and perceptions of employees across different hierarchical levels. A tailored communication strategy is critical to ensure that artificial intelligence adoption aligns with the strategic and operational goals of the organization while fostering employee acceptance and engagement (Morosan and Dursun-Cengizci, 2024).

For nonmanagerial employees, communication should focus on the tangible benefits of artificial intelligence in daily tasks. Emphasizing how artificial intelligence can streamline processes, reduce operational costs and handle repetitive tasks more efficiently can help these employees see the practical advantages of technology adoption. Demonstrating real-world examples, such as using artificial intelligence-driven chatbots to improve task efficiency, can transform perceived threats into opportunities for professional growth and job satisfaction.

For first-line managers, the messaging should highlight artificial intelligence’s capacity to enhance service personalization and improve customer experiences. By showcasing how artificial intelligence can analyze customer data and predict preferences, first-line managers can better understand its role in fostering customer loyalty and boosting service quality. This approach aligns with recent findings that service personalization through artificial intelligence can significantly enhance customer satisfaction (Ersoy and Ehtiyar, 2023).

For top managers, communication strategies should focus on the strategic and competitive advantages of adopting artificial intelligence. Highlighting how early adoption can lead to operational efficiencies, improved profitability and strengthened market positioning can motivate top executives to champion artificial intelligence initiatives. This aligns with studies suggesting that a clear link between technological investments and competitive advantage strengthens managerial commitment (Bulchand-Gidumal et al., 2024).

Addressing barriers to AI adoption, such as resistance to change and limited awareness, is equally important. Resistance can be mitigated through transparent communication, active involvement of employees in decision-making processes and comprehensive training programs that build confidence and competence in using artificial intelligence tools (Li et al., 2022). Educational initiatives, workshops and hands-on experiences can bridge knowledge gaps and foster a culture of curiosity and openness. In addition, addressing interorganizational concerns such as privacy and security is critical. These challenges can be managed through the implementation of robust data protection measures ensuring compliance with regulations like GDPR and transparent communication regarding data usage and security protocols.

To facilitate sustained artificial intelligence integration, hospitality organizations should prioritize targeted training programs focusing on technological adaptability, customer service enhancement and innovation management. Collaborating with educational institutions can help attract talent with a proactive attitude toward technological advancements. Offering career development opportunities that align with artificial intelligence-driven roles can also support employee retention and skill development.

All in all, a hierarchical-specific approach to artificial intelligence adoption can nurture a positive organizational culture: By aligning strategic objectives with tailored communication mechanisms, hospitality managers can drive both technological and human-centric excellence, reinforcing the industry’s adaptability in an increasingly digital landscape.

While our study provides valuable insights into artificial intelligence adoption within the hospitality sector, it is important to acknowledge its limitations. One significant limitation of this study is its case study-oriented design, which focused on participants primarily from the hotel industry in a specific geographical location. This narrow scope may constrain the generalizability of the findings to other segments within the broader tourism industry. To address this, it would be beneficial to consider the application of additional methods (e.g., Delphi Method) for a more in-depth investigation into the mutual influence of the identified motivators and obstacles. Moreover, consensus mapping could be used across different geographical regions to provide an overview of perceptions within varying cultural contexts. Conducting the same study in different countries would help account for cultural differences and significantly enhance the generalizability of the findings. In addition, our study focused solely on perceptions within hierarchical positions in hotels, potentially overlooking valuable insights from other stakeholders, such as customers or AI technology providers.

To address these limitations and expand the scope of future research, several promising avenues emerge. First, future studies could explore AI adoption across various segments of the broader tourism industry, including airlines, cruise lines and travel agencies. This broader perspective would provide a more comprehensive understanding of the challenges and opportunities associated with AI adoption within different sectors of the tourism industry. Furthermore, future research could investigate the perceptions and attitudes of other stakeholders beyond employees, such as suppliers, AI technology providers and regulatory bodies. Understanding the perspectives of these diverse stakeholders could shed light on additional barriers and facilitators to AI adoption and inform more holistic strategies for AI implementation.

In addition, exploring the potential applications of AI beyond traditional hospitality settings could offer new insights and opportunities for innovation. For example, research could examine the use of AI in ecotourism initiatives, destination management or cultural heritage preservation within the tourism industry. By broadening the scope of inquiry to include diverse industries and applications, future research can contribute to a more nuanced understanding of AI adoption dynamics and facilitate the development of tailored strategies for successful implementation of AI solutions across and beyond the tourism sector.