How conversational AI reshapes tourist decision-making systems: Extending the S-O-R framework through emotional and contextual mediators

Abstract

Purpose

This study investigates how conversational AI, particularly ChatGPT, influences tourists’ purchase intention by reshaping decision-making processes within tourism service systems. It contributes to a deeper understanding of human–AI interaction and its role in contemporary consumer decision systems in the tourism and hospitality context. The study focuses on the mediating roles of perceived empathy, trust in AI-generated content, and content adaptability—three complementary emotional, cognitive, and contextual mechanisms that are integrated within a unified framework to explain how individuals engage with AI during travel planning and decision-making. Grounded in the Stimulus–Organism–Response (S-O-R) framework, the research captures both cognitive and emotional pathways linking AI interaction to behavioral outcomes.

Design/methodology/approach

Out of 2,256 responses, 1,438 tourists who confirmed using ChatGPT for travel-related decisions were retained for analysis.

Findings

The findings reveal that conversational AI significantly and directly influences tourists’ purchase intention, while also enhancing users’ perceived empathy during AI interaction, trust, and adaptability. These psychological responses partially mediate the relationship, highlighting the dual role of emotional and contextual mechanisms in AI-driven decision-making systems.

Originality/value

The study extends the S-O-R framework by simultaneously integrating emotional (perceived empathy), cognitive (trust in AI-generated content), and contextual (content adaptability) dimensions within a unified AI-enabled decision-making framework. From a managerial perspective, the findings emphasize the strategic importance of designing AI systems that foster trust, simulate empathy, and dynamically adapt to user preferences. Such integrated capabilities support more effective, self-directed decision-making processes and enhance engagement and conversion outcomes in tourism service ecosystems.

Keywords

ChatGPT perceived empathy trust in AI-generated content content adaptability purchase intention

Introduction

Tourist purchase intention has long been a central concern in tourism and hospitality research due to its direct implications for market performance, service design, and strategic destination management. A wide range of factors has been identified as influencing this intention, including destination image, price perception, service quality, and psychological determinants such as motivation, perceived risk, and personal involvement.¹ In recent years, digital transformation has introduced a new category of influential variables driven by artificial intelligence (AI) and interactive technologies.² As tourists increasingly rely on digital tools throughout their travel journey, the psychological and experiential dimensions of technology-mediated interactions have become critical in shaping purchasing behavior.³

Among the most notable innovations in this domain is ChatGPT, a generative AI model capable of engaging users in dynamic, human-like conversational interactions.⁴ Unlike traditional travel websites or static recommendation systems, ChatGPT allows users to interact in real time, refine their queries, and receive responses that are coherent, contextually relevant, and tailored to individual needs.⁵ This interactive capability positions ChatGPT not merely as an information provider, but as a potential influencer of consumer behavior, particularly in shaping tourist purchase intention.⁶

The growing role of ChatGPT in tourism and hospitality reflects broader shifts in how tourists seek, process, and respond to information.⁷ As digital travelers increasingly demand speed, accuracy, and personalization, ChatGPT offers a distinct advantage by delivering intelligent content through an intuitive conversational interface.⁸ This functionality enables users to explore destinations, accommodations, and travel logistics while receiving immediate, tailored responses, thereby facilitating smoother transitions from information search to decision-making and enhancing purchase intention.⁹

Beyond its informational role, ChatGPT influences key psychological drivers of tourist behavior, particularly perceived empathy, trust in AI-generated content, and content adaptability.^10,11 In tourism contexts—where decisions are often emotionally driven—these factors play a critical role in shaping behavioral outcomes. When users perceive the system as responding in ways that reflect an understanding of their needs, providing reliable information, and adapting to their preferences, they are more likely to develop confidence and intention to act.¹² Accordingly, the effectiveness of ChatGPT lies not only in the information it provides but in the quality of interaction it delivers, combining perceived emotional responsiveness, cognitive assurance, and contextual relevance.

This study is conceptually grounded in the Stimulus–Organism–Response (S-O-R) framework, which provides a robust lens for understanding how external stimuli influence internal cognitive and emotional states, ultimately leading to behavioral responses.¹³ Within this framework, ChatGPT represents the stimulus, triggering organism-level processes such as perceived empathy, trust in AI-generated content, and content adaptability, which in turn shape the response, namely tourist purchase intention.¹⁴ This perspective enables a more nuanced understanding of how AI-mediated interactions translate into behavioral outcomes in tourism and hospitality settings.

Despite the increasing adoption of AI in tourism and hospitality,¹⁵ existing research remains limited in capturing the experiential and behavioral implications of AI-driven conversational systems such as ChatGPT. Prior studies have predominantly focused on functional aspects of AI, including automation, efficiency, and algorithmic personalization,^16–18 while largely overlooking the psychological and emotional dynamics perceived within human–AI interactions. Moreover, although constructs such as trust, empathy, and perceived relevance are well established in consumer behavior research,^14,19 limited research has examined their collective and integrated role within a unified AI-mediated conversational framework, particularly in tourism contexts. Existing studies tend to examine these constructs in isolation or within static digital environments, failing to capture their dynamic interplay in real-time, adaptive interactions. In addition, the application of established behavioral frameworks, such as the S-O-R model, to generative AI contexts remains limited, particularly in tourism settings. This highlights a critical need for empirical research that explains how conversational AI stimuli translate into internal psychological states and ultimately influence tourist behavior.

Addressing these gaps, this study aims to provide an integrated understanding of how ChatGPT influences tourist purchase intention through the combined effects of emotional, cognitive, and contextual mechanisms within a single conceptual framework. Specifically, the study seeks to: 1) examine the effect of ChatGPT on purchase intention, perceived empathy, trust in AI-generated content, and content adaptability; 2) assess the impact of perceived empathy, trust, and content adaptability on purchase intention; and 3) investigate the mediating roles of these psychological factors in the relationship between ChatGPT and purchase intention. By adopting the Stimulus–Organism–Response (S-O-R) framework, this research not only extends consumer behavior theory into the emerging domain of generative AI in tourism, but also advances the understanding of how dynamic, conversational AI systems reshape decision-making processes in experience-driven industries. Furthermore, the study contributes to the literature by simultaneously integrating emotional (perceived empathy), cognitive (trust in AI-generated content), and contextual (content adaptability) constructs within a real-time interactive context, thereby offering a more holistic and theoretically integrated explanation of tourist behavior in digital environments. From a practical perspective, the findings provide actionable insights for tourism practitioners and destination managers on how to strategically leverage conversational AI technologies to enhance user engagement, build trust, personalize interactions, and ultimately stimulate purchase intentions in increasingly competitive digital marketplaces.

Literature review and hypothesis development

The increasing adoption of conversational artificial intelligence in tourism and hospitality has fundamentally reshaped how travelers search for information, evaluate alternatives, and formulate purchasing decisions.²⁰ Among recent AI applications, ChatGPT has attracted considerable scholarly and practical attention because of its ability to generate interactive, context-sensitive, and human-like communication in real time.^21,22 Unlike traditional digital systems that primarily provide static information or standardized recommendations, ChatGPT enables continuous conversational exchanges through which users can clarify preferences, refine travel needs, and receive dynamically tailored responses.⁹ In tourism and hospitality settings, purchasing decisions are shaped not only by the quality of information provided, but also by the extent to which communication appears responsive, personalized, and emotionally reassuring.^22,23

To explain these interaction dynamics, the present study adopts the Stimulus–Organism–Response (S-O-R) framework, which proposes that external environmental stimuli influence individuals’ internal psychological states, which subsequently shape behavioral responses.^24,25 Within the context of conversational AI, interaction with ChatGPT represents the external stimulus because users are exposed to AI-generated communication characterized by personalization, responsiveness, contextual adaptation, and conversational engagement.^26,27 These interaction features may activate multiple organism-level psychological evaluations that shape how tourists interpret and respond to the communication experience.²⁸

More specifically, the present study conceptualizes perceived empathy, trust in AI-generated content, and content adaptability as complementary organism-level mechanisms that explain how tourists psychologically process interaction with conversational AI.^27,28 Perceived empathy reflects the emotional dimension of the interaction, capturing the extent to which users perceive the system as attentive, understanding, and responsive to their needs.²⁶ Trust in AI-generated content represents a cognitive evaluation concerning the credibility, reliability, and dependability of the information generated by ChatGPT.²⁹ Content adaptability, in turn, reflects a contextual evaluation related to the system’s ability to dynamically tailor communication according to users’ preferences, situational conditions, and evolving travel requirements.^30–32 Collectively, these mechanisms provide a multidimensional interpretation of the organism stage by integrating emotional, cognitive, and contextual responses within a unified AI-mediated tourism framework.

Drawing on the S-O-R perspective, these internal psychological responses are expected to shape tourists’ behavioral outcomes, particularly their intention to purchase tourism and hospitality services. Accordingly, the following sections develop the study hypotheses by examining the direct effects of interaction with ChatGPT on perceived empathy, trust in AI-generated content, content adaptability, and purchase intention, as well as the mediating roles of these organism-level mechanisms in explaining AI-influenced tourist behavior.

ChatGPT and intention to purchase

Conversational AI systems increasingly influence consumer decision-making by facilitating interactive and personalized experiences.²⁰ ChatGPT enables travelers to obtain immediate recommendations, refine preferences, and receive tailored responses aligned with their situational needs.²¹ Such functionality may reduce decision complexity and reinforce confidence throughout the travel planning process.⁹ Prior research has shown that users are more likely to develop favorable behavioral intentions when digital interactions are perceived as useful, responsive, and contextually relevant.²² Given the experiential nature of tourism consumption, real-time conversational support may further encourage users to rely on AI-generated recommendations when making purchase decisions.²⁵ Accordingly, the following hypothesis is developed:

H1: Interaction with ChatGPT has a significant positive effect on tourists’ intention to purchase tourism and hospitality services.

ChatGPT and perceived empathy

Perceived empathy has become an increasingly important aspect of digital communication quality, particularly in interactions involving personalized assistance and emotional engagement.²⁶ ChatGPT can generate responses that appear attentive, supportive, and aligned with users’ concerns and preferences 6. This responsiveness may create a perception of being understood and acknowledged during the interaction process, thereby supporting emotional engagement with the system.¹⁴ Studies examining AI-mediated communication suggest that systems capable of simulating empathetic interaction can enhance perceptions of supportiveness and relational closeness.²⁷ This perceived responsiveness may play a particularly important role during tourism-related decision evaluation, where travelers often experience uncertainty and actively seek reliable guidance.³³ So, the following hypothesis is put forward:

H2: Interaction with ChatGPT significantly enhances tourists’ perception of empathy in the digital communication experience.

ChatGPT and trust in AI-generated content

Trust remains a critical factor in digital tourism environments because travelers frequently depend on online information when making financially and experientially significant decisions.²⁹ As a generative AI application, ChatGPT produces detailed and contextually tailored recommendations that may influence users’ evaluations of reliability and credibility.^30,31 Unlike static recommendation systems, conversational AI delivers adaptive responses that evolve according to user input and situational context.^25,32 Previous research indicates that trust in AI-generated communication is influenced by informational clarity, coherence, usefulness, and perceived neutrality.^15,34,35 When tourists perceive AI-generated content as dependable and accurate, they are more likely to rely on it during decision-making processes.^36,37 Hence, the following hypothesis is proposed:

H3: Interaction with ChatGPT positively affects tourists’ trust in AI-generated content.

ChatGPT and content adaptability

A distinguishing feature of advanced conversational AI systems is their ability to adapt communication dynamically according to user input and evolving interaction contexts.^4,38 ChatGPT demonstrates this adaptability by modifying response structure, tone, detail, and thematic focus in real time.³⁹ Such flexibility is particularly valuable in tourism and hospitality because travelers often seek recommendations that reflect individual motivations, constraints, and travel preferences.⁴⁰ Context-sensitive communication can improve perceived usefulness and enhance overall interaction quality.⁴¹ When users perceive that the system adjusts effectively to their changing needs, they may evaluate the communication experience as more relevant and personalized.^42,43 Accordingly, the following hypothesis is articulated:

H4: Interaction with ChatGPT significantly enhances tourists’ perception of content adaptability.

Perceived empathy and intention to purchase

Empathetic communication has been widely associated with stronger engagement and more favorable behavioral outcomes in digital environments.⁴⁴ Tourism-related decisions frequently involve emotional uncertainty, anticipation, and experiential expectations, making emotionally supportive interaction especially important.^23,45 When users perceive that ChatGPT responds in a thoughtful and attentive manner, they may feel more psychologically comfortable and emotionally connected during the interaction process.¹⁴ Such perceptions can reduce hesitation and support confidence in acting upon the recommendations provided by the system.⁴⁶ Therefore, the following hypothesis is proposed:

H5: Tourists’ perception of empathy during their interaction with ChatGPT positively influences their intention to purchase tourism and hospitality services.

Trust in AI-generated content and intention to purchase

Behavioral intention in digital environments is strongly influenced by the degree to which users trust the available information and recommendations.⁴⁷ In tourism and hospitality, purchasing decisions often involve uncertainty because consumers must evaluate intangible experiences before consumption occurs.¹⁵ AI-generated content that is perceived as credible, reliable, and useful may therefore strengthen users’ willingness to rely on technological recommendations.⁴⁸ Prior studies suggest that trust reduces perceived risk and increases confidence in decision-making processes, particularly within technology-mediated service contexts.⁴⁹ As a result, tourists who trust AI-generated travel recommendations may become more willing to translate their evaluations into actual purchasing intentions.^50,51 Accordingly, the following hypothesis is formulated:

H6: Tourists’ trust in AI-generated content positively influences their intention to purchase tourism and hospitality services.

Content adaptability and intention to purchase

Personalized communication has become increasingly important in digitally mediated tourism experiences because travelers expect information that aligns with their specific preferences and situational needs.⁵ ChatGPT allows interactions to evolve dynamically according to users’ questions, concerns, and travel objectives.¹¹ This adaptability may increase users’ perceptions of relevance and interaction quality, particularly when recommendations appear tailored to individual circumstances.⁶ Greater alignment between user expectations and AI-generated communication can enhance engagement and improve confidence in decision-making.²⁷ Consequently, adaptable communication may increase users’ willingness to rely on AI-supported recommendations when considering tourism-related purchases.^52,53 Based on this rationale, the following hypothesis is proposed:

H7: Tourists’ perception of content adaptability positively influences their intention to purchase tourism and hospitality services.

Perceived empathy as a mediator

The effectiveness of conversational AI may depend not only on informational quality but also on the emotional experience generated during interaction.^54,55 Perceived empathy represents an affective mechanism through which users evaluate whether the system appears attentive and responsive to their needs.⁴⁴ AI systems capable of simulating empathetic communication may encourage users’ feelings of reassurance, relational closeness, and emotional comfort.³⁰ Such emotional responses are particularly relevant in tourism planning situations where travelers seek both informational guidance and psychological reassurance.¹⁴ Through this process, perceived empathy may help explain how interaction with ChatGPT translates into stronger purchasing intentions.^56,57 So, the following hypothesis is advanced:

H8: Perceived empathy mediates the relationship between interaction with ChatGPT and tourists’ intention to purchase.

Trust in AI-generated content as a mediator

The influence of conversational AI on behavioral outcomes is likely to depend on how users evaluate the credibility and reliability of AI-generated information.³⁴ Although ChatGPT can provide detailed and persuasive recommendations, purchasing intention may not emerge unless users perceive the generated content as trustworthy.³⁵ Trust therefore functions as an important evaluative mechanism through which users determine whether AI-generated communication is dependable enough to support decision-making.³⁶ Prior research suggests that higher levels of trust increase users’ confidence in acting upon technological recommendations and reduce uncertainty during online decision processes.⁵⁸ Accordingly, trust in AI-generated content may serve as an important pathway linking interaction with ChatGPT to tourism-related purchasing intention. Hence, the following hypothesis is proposed:

H9: Trust in AI-generated content mediates the relationship between interaction with ChatGPT and tourists’ intention to purchase.

Content adaptability as a mediator

The ability of conversational AI to adjust communication dynamically may influence behavioral intention indirectly through users’ perceptions of adaptability and relevance.⁶ ChatGPT can modify responses according to evolving user preferences, situational constraints, and interaction patterns, thereby creating a more personalized communication experience.⁵⁹ Such flexibility may strengthen perceptions of usefulness and increase engagement with the interaction process.²⁷ When travelers perceive that AI-generated recommendations align closely with their needs and planning logic, they may become more receptive to the system’s suggestions.³⁸ Through this mechanism, content adaptability may help transform interactive communication into stronger tourism-related purchasing intentions.⁹ So, the following hypothesis is presented:

H10: Perceived content adaptability mediates the relationship between interaction with ChatGPT and tourists’ intention to purchase.

Figure 1 illustrates the proposed research model, which outlines the relationships between ChatGPT and tourists’ intention to purchase, mediated by perceived empathy, trust in AI-generated content, and content adaptability, based on the S-O-R theoretical framework.

Figure 1.

Proposed research model

Measurement items were adapted from Han et al.¹⁶ for ChatGPT, Shalan¹¹ for perceived empathy, Pham et al.¹⁷ for trust in AI-generated content, Shalan et al.⁵⁴ for content adaptability, and Pillai et al.¹⁸ for purchase intention.

Methods

Sample and data 5collection

This study employed an online survey methodology to investigate tourists’ behavioral responses to their interac ions with ChatGPT during the travel planning process. Given the digital nature of both the research topic and the target population, online data collection was deemed the most appropriate and efficient approach for capturing relevant user experiences. The survey was designed to reach individuals who had specifically used ChatGPT to support their travel-related decision-making, such as comparing destinations, planning itineraries, or seeking real-time travel guidance. A non-probability convenience sampling technique was adopted to access this population through social media platforms where tourism discussions and technology use are common. The survey link was distributed via travel-focused communities and pages on Facebook, Instagram, TikTok, and LinkedIn, capitalizing on the platforms’ broad user reach and the high likelihood of encountering tech-savvy travelers within these environments. The distribution campaign included public posts, direct messages, and promoted content targeting users who had demonstrated interest in both travel and digital tools.

Data collection took place over a 10-week period between April and mid-June 2025, using a structured electronic questionnaire hosted online. The questionnaire began with a short preamble outlining the study’s purpose, providing a clear explanation of ChatGPT’s relevance to the research, and assuring respondents of the confidentiality of their responses. Consent was obtained electronically before proceeding with the questions. To ensure that only relevant participants were included in the final analysis, two eligi bility screening questions were integrated at the beginning of the survey. The first asked whether the individual had used ChatGPT within the past year for tourism-related inquiries. Respondents who answered “no” were disqualified. Those who answered “yes” were then asked whether their usage was limited to general exploration or involved actual decision-making—such as selecting destinations or planning activities. Only participants who reported decision-support usage were allowed to proceed.

A total of 2,256 individuals accessed the survey. Among them, 1,438 respondents met the inclusion criteria and completed the questionnaire, while the remaining 818 participants were excluded based on their responses to the screening items. The final dataset of 1,438 valid responses constituted the core sample for the study’s empirical analyses and hypothesis testing.

Measures

To empirically examine the constructs included in the proposed model, the study utilized a structured survey instrument comprising validated multi-item scales adapted from prior research. All measurement items were rated on a five-point Likert scale. The independent variable, ChatGPT usage, was assessed using a three-item scale adapted from Han et al.¹⁶ A representative item includes: “I will continue using ChatGPT for hospitality/tourism information searches.” To evaluate perceived empathy, six items were adapted from Shalan.¹¹ A sample statement is: “ChatGPT communicates with empathy and shows understanding of my travel-related concerns.” Trust in AI-generated content was measured using a four-item scale drawn from Pham et al.¹⁷ For example: “ChatGPT’s response and advice can meet my expectations.” The construct content adaptability was operationalized through five items sourced from Shalan et al.⁵⁴ An illustrative item is: “ChatGPT customizes its recommendations based on my individual travel preferences.” The dependent variable, intention to purchase, was measured using a four-item scale adapted from Pillai et al.¹⁸ A representative item reads: “I would like to repeat my experience in digital assistants.” In addition to the main constructs, the survey collected basic demographic data, including gender, age, nationality, and educational background.

Common method biases

Given the reliance on self-reported data collected through a single measurement instrument in a cross-sectional design, the study proactively addressed the potential issue of common method bias (CMB), which may compromise the integrity of inter-variable relationships by introducing artificial inflation or deflation. This concern is especially pertinent when all constructs are measured simultaneously using perceptual ratings from the same respondents. To evaluate the presence and extent of CMB, two diagnostic techniques commonly cited in behavioral research were employed: Harman’s one-factor test and Principal Component Analysis (PCA), following the recommendations of Hair et al.⁶⁰ All measurement items were subjected to an unrotated exploratory factor analysis. The results indicated that no single factor accounted for a substantial portion of the total variance, and the first emerging factor explained less than 50% of the overall variance structure. These findings suggest that the likelihood of systematic bias arising from the data collection method is limited. Therefore, concerns related to common method variance are not expected to significantly distort the relationships tested in this study, and the measurements can be considered reasonably robust for the purposes of subsequent statistical analysis.

Results

Demographics profile

Table 1 presents the demographic characteristics of the study participants. The gender distribution was nearly equal, with males representing 50.1% (n = 721) and females accounting for 49.9% (n = 717). Regarding age, the majority of respondents fell within the 21 to 30 age group (45.8%), followed by those aged 31 to 40 (33.9%). Participants under 21 years constituted 2.2% of the sample, while individuals aged 41 to 50 and those over 50 represented 9.2% and 8.9%, respectively. In terms of educational attainment, most respondents held a bachelor’s degree (65.9%), while 23.8% had qualifications below the bachelor level, and 10.3% reported having postgraduate education. As for nationality, the sample included 46.9% Arab participants and 53.1% non-Arab (foreign) nationals.

Table 1.

Demographics profile.

Measure	Item	Frequencies	Percentage
Gender	Male	721	50.1%
Gender	Female	717	49.9%
Age	Under 21 years	32	2.2%
	21 to 30 years	658	45.8%
	31 to 40 years	488	33.9%
	41 to 50 years	132	9.2%
	50 years and above	128	8.9%
Educational level	Less than bachelor	342	23.8%
	Bachelor	948	65.9%
	Postgraduate	148	10.3%
Nationality	Arab nationality	674	46.9%
Nationality	Foreign nationality	764	53.1%

Source: Authors’ own work

Measurement model

To validate the constructs prior to testing structural relationships, a rigorous examination of the measurement model was conducted. This step was essential to ensure that each construct was accurately represented and measured with both consistency and conceptual alignment. The analysis of factor loadings confirmed that all items displayed strong associations with their respective constructs. Every item exceeded the minimum threshold of 0.70, with values ranging from 0.834 to 0.903, indicating that the observed variables reliably captured the dimensions they were intended to measure. Reliability across constructs was demonstrated through Cronbach’s alpha coefficients, all of which were above 0.87, as well as composite reliability (CR) scores surpassing 0.91. These indicators reflect a high degree of internal consistency, suggesting that each scale was cohesive and dependable for capturing participant responses. In terms of convergent validity, each latent construct achieved an average variance extracted (AVE) greater than 0.70. This means that the majority of variance in the observed indicators was explained by their corresponding latent variable—thereby confirming that the items effectively converged to represent single theoretical concepts. To rule out concerns of multicollinearity, variance inflation factor (VIF) values were also examined. All constructs reported VIF scores well below the standard cutoff of 5, ranging between 2.378 and 3.112, confirming that redundancy between predictors was not an issue in this model (Table 2).

Table 2.

Measurement model.

Construct	Loading	Cronbach’s alpha	CR	AVE	VIF
ChatGPT		0.884	0.919	0.749	2.675
ChatGPT1	0.898
ChatGPT2	0.872
ChatGPT3	0.901
Perceived empathy		0.810	0.950	0.762	3.112
Perceived empathy1	0.887
Perceived empathy2	0.899
Perceived empathy3	0.870
Perceived empathy4	0.888
Perceived empathy5	0.834
Perceived empathy6	0.858
Trust in AI-generated content		0.943	0.926	0.759	2.547
Trust in AI-generated content1	0.867
Trust in AI-generated content2	0.895
Trust in AI-generated content3	0.886
Trust in AI-generated content4	0.834
Content adaptability		0.901	0.942	0.766	2.378
Content adaptability1	0.857
Content adaptability2	0.902
Content adaptability3	0.889
Content adaptability4	0.856
Content adaptability5	0.873
Intention to purchase		0.877	0.933	0.779	2.555
Intention to purchase1	0.903
Intention to purchase2	0.897
Intention to purchase3	0.881
Intention to purchase4	0.850

Source: Authors’ own work

As part of validating the measurement model, it was essential to assess the degree to which the latent constructs in the study were empirically distinct from one another. Discriminant validity ensures that each construct captures a unique aspect of the theoretical framework and does not substantially overlap with other constructs in the model. Two widely recognized techniques were employed to verify discriminant validity: the Fornell–Larcker criterion and the Heterotrait-Monotrait (HTMT) ratio.⁶⁰ The Fornell–Larcker approach involves comparing the square root of the average variance extracted (AVE) for each construct with the correlations it shares with other constructs. Table 3 displays the square roots of AVE along the diagonal and the inter-construct correlations in the lower triangle. The values on the diagonal were all higher than the corresponding off-diagonal correlations. For example, the AVE square root for ChatGPT was 0.890, which exceeded its correlations with Perceived Empathy (0.564), Trust in AI-generated Content (0.609), Content Adaptability (0.568), and Intention to Purchase (0.666). This pattern was consistent across all constructs, confirming that each construct shared more variance with its own indicators than with any other variable in the model. Accordingly, the Fornell–Larcker criterion was satisfied, indicating that discriminant validity was established.

Table 3.

Squared roots of AVE.

	ChatGPT	Perceived empathy	Trust in AI-generated content	Content adaptability	Intention to purchase
ChatGPT	0.890
Perceived empathy	0.564	0.872
Trust in AI-generated content	0.609	0.459	0.870
Content adaptability	0.568	0.509	0.511	0.875
Intention to purchase	0.666	0.601	0.532	0.564	0.882

Source: Authors’ own work

To strengthen this conclusion, the Heterotrait-Monotrait (HTMT) ratio was also examined. HTMT is considered a more rigorous method for evaluating discriminant validity, particularly when constructs are conceptually related. Table 4 presents the HTMT values, all of which fell comfortably below the conservative threshold of 0.85. For instance, the HTMT ratio between Content Adaptability and Intention to Purchase was 0.549, while the ratio between Trust in AI-generated Content and Perceived Empathy was 0.454. These moderate values suggest that while the constructs are theoretically connected, they remain empirically distinct and not redundant.

Table 4.

Heterotrait Monotrait (HTMT) ratio.

	ChatGPT	Perceived empathy	Trust in AI-generated content	Content adaptability
ChatGPT
Perceived empathy	0.447
Trust in AI-generated content	0.501	0.454
Content adaptability	0.432	0.550	0.499
Intention to purchase	0.509	0.384	0.532	0.549

Source: Authors’ own work

Structure model

The structural model was tested to examine both direct and mediating effects among the study’s key constructs. The results, as shown in Table 5, provide strong empirical support for all proposed hypotheses, affirming the validity of the conceptual framework and the suitability of ChatGPT as a meaningful predictor of tourist behavior in digital environments. In terms of direct relationships, H1 was supported, showing that ChatGPT has a significant and positive influence on intention to purchase (β = 0.532, p-value >0.000). This finding confirms that interaction with AI-powered systems plays a key role in shaping consumer behavior in the context of tourism and hospitality. Furthermore, ChatGPT was found to significantly predict perceived empathy (H2: β = 0.378, p-value >0.000), trust in AI-generated content (H3: β = 0.447, p-value >0.000), and content adaptability (H4: β = 0.501, p-value >0.000). These outcomes indicate that users recognize ChatGPT as not only a functional source of information but also an adaptive system capable of generating responses perceived as contextually and emotionally responsive. The results also confirmed the significance of the perceptual and emotional constructs in influencing behavioral intention. Perceived empathy had a strong positive effect on intention to purchase (H5: β = 0.428, p-value >0.000), as did trust in AI-generated content (H6: β = 0.398, p-value >0.000) and content adaptability (H7: β = 0.408, p-value >0.000). These results highlight the central role of user perceptions— particularly users’ perceptions that the system responds appropriately and adapts effectively to their needs—in translating AI interaction into consumer commitment.

Table 5.

Structure model.

Path	β	s.e	C.R	p-value	Results
Direct path
H1: ChatGPT → Intention to purchase	0.532	0.056	9.502	0.000	Accepted
H2: ChatGPT → Perceived empathy	0.378	0.060	6.300	0.000	Accepted
H3: ChatGPT → Trust in AI-generated content	0.447	0.063	7.095	0.000	Accepted
H4: ChatGPT → Content adaptability	0.501	0.059	8.491	0.000	Accepted
H5: Perceived empathy → Intention to purchase	0.428	0.068	6.294	0.000	Accepted
H6: Trust in AI-generated content → Intention to purchase	0.398	0.050	7.962	0.000	Accepted
H7: Content adaptability → Intention to purchase	0.408	0.072	5.66	0.000	Accepted
Mediation path
H8: ChatGPT → Perceived empathy → Intention to purchase	0.161	0.049	3.285	0.000	Accepted
H9: ChatGPT → Trust in AI-generated Content → Intention to purchase	0.177	0.053	3.339	0.000	Accepted
H10: ChatGPT → Content adaptability → Intention to purchase	0.204	0.045	4.533	0.000	Accepted

Source: Authors’ own work

The results further demonstrated that ChatGPT’s impact on tourists’ intention to purchase is exerted through both direct and indirect pathways. In addition to the significant direct effect, the model revealed three statistically significant indirect effects via mediating variables. Perceived empathy mediated the relationship between ChatGPT and intention to purchase (H8: β = 0.161, p-value = 0.000), indicating that interactions perceived as emotionally responsive play a role in shaping behavioral intentions. Similarly, trust in AI-generated content was found to function as a mediator (H9: β = 0.177, p-value >0.000), reinforcing the importance of content credibility and user confidence in driving consumer action. The most pronounced mediation effect was observed through content adaptability (H10: β = 0.204, p-value >0.000), highlighting the critical influence of dynamic, personalized responses on tourist decision-making. However, since the direct path between ChatGPT and intention to purchase (H1: β = 0.532, p-value >0.000) remained statistically significant even after accounting for these mediators, the results clearly indicate that all three mediation effects are partial rather than full. This suggests that while perceived empathy, trust, and adaptability explain a meaningful portion of ChatGPT’s influence, the system also maintains a direct behavioral impact on its own.

Discussion

This study examined how tourists’ interactions with ChatGPT influence their intention to purchase tourism and hospitality services, both directly and through key psychological mediators. Grounded in the S-O-R framework, the findings illuminate the complex interplay between technological stimuli and internal tourist responses that ultimately shape behavioral outcomes. The results contribute to the growing discourse on AI adoption in tourism by providing empirical evidence on the cognitive and emotional pathways through which conversational AI affects decision-making. The analysis confirmed that ChatGPT has a strong direct effect on tourists’ intention to purchase, highlighting its potential as a technology-driven influence on tourist behavior rather than merely an information tool. Shi et al.¹² highlight that when tourists engage with ChatGPT, the system’s real-time responsiveness, structured interaction, and ability to provide tailored travel recommendations contribute to a heightened sense of confidence and clarity. Farahat⁹ further argues that the simulation of responsive dialogue in AI systems mirrors aspects of human travel consultancy, prompting users to perceive such tools as credible, practical, and worthy of reliance for decision-making. In line with these perspectives, the present study confirms that ChatGPT functions not merely as a source of information, but as a technological mechanism capable of shaping tourist behavior through interactive communication.

Furthermore, the findings demonstrated that ChatGPT significantly enhances perceived empathy among tourists. Gao and Zhang⁶¹ assert that in emotionally driven travel contexts—such as selecting a sentimental destination, planning family vacations, or seeking reassurance during solo journeys—tourists are more receptive to systems that appear responsive to their emotional cues. Kodal⁶² and Sousa et al.⁶³ clarify that empathy in this setting is not about genuine emotional intelligence, but rather about the tourist’s personal perception that the system is able to “understand” their needs and concerns. Ferreira and Pereira⁶⁴ and Ferrer-Rosell et al.³⁶ emphasize that this perceived emotional resonance fosters a perceived sense of relational closeness and sustained engagement, both of which are critical in transforming digital interactions into actual purchasing behavior.

The study also revealed that ChatGPT positively affects tourists’ trust in AI-generated content. Al-Romeedy and Singh² emphasize that trust is a fundamental element in digital interactions, especially in contexts where human support is absent and decisions involve considerable financial or temporal commitment. Stergiou and Nella⁵ explain that when tourists perceive the information provided by ChatGPT as accurate, impartial, and personalized, they are more inclined to incorporate its recommendations into their travel planning. Xinlin et al.³⁸ further argue that trust serves as the cognitive anchor of the tourist’s decision-making journey, reducing uncertainty and enhancing their willingness to depend on AI-driven systems.

In addition, the results indicated that ChatGPT significantly improves perceptions of content adaptability. Stergiou and Nella⁵ and Shalan et al.⁵⁴ argue that unlike static information sources or rigid booking interfaces, ChatGPT facilitates a dynamic, conversational exchange in which tourists can adjust their preferences, consider alternatives, and receive tailored suggestions. Huang et al.⁶⁵ note that this interactive flexibility fosters the impression that the system responds appropriately to the tourist’s specific goals, limitations, and situational needs. Morrison¹ and Leveau¹⁵ further suggest that as perceived adaptability increases, so does the tool’s practical value, leading to greater user satisfaction and a stronger intention to act upon the recommendations provided.

The model further confirmed that perceived empathy has a significant positive effect on tourists’ intention to purchase. Li and Lee⁶⁶ indicate that when tourists perceive themselves as emotionally acknowledged during their interaction with ChatGPT, they are more inclined to trust the system, feel guided in their planning process, and follow its recommendations. Zaki and Al-Romeedy¹⁴ and Dixit²⁷ emphasize that emotion-based constructs are equally—if not more—important than purely informational elements in influencing consumer behavior within digital tourism contexts. Their findings underscore that anticipation, emotional resonance, and personal relevance are central to decision-making in tourism, a sector deeply rooted in experiential and affective value.

Similarly, the findings highlighted that trust in AI-generated content significantly contributes to intention to purchase. Balaskas et al.⁶⁷ emphasize that trust plays a dual role in digital decision-making—not only by reducing perceived risk, but also by streamlining the process and encouraging users to take action. Foroughi et al.¹⁰ argue that in situations characterized by information overload or contradictory online reviews, a trusted AI system like ChatGPT can serve as a useful source of clarity and direction. Ramos and Ramos⁶⁸ further note that when tourists view the system’s content as coherent, objective, and logically sound, they are more likely to transition confidently from exploration to purchase, minimizing hesitation and maximizing decision efficiency.

The results also established that content adaptability significantly enhances intention to purchase. Ilieva et al.⁶⁹ highlight that when tourists receive travel suggestions that align with their specific interests—be they cultural, financial, or logistical—they are more likely to perceive the interaction as meaningful and are thus more inclined to act on it. Singh et al.⁷⁰ add that this form of personalization fosters a perception of individualized attention within the digital environment, which not only strengthens trust in the system but also enhances the tourist’s perceived autonomy and control during the planning process.

Finally, the study demonstrated that perceived empathy, trust in AI-generated content, and content adaptability all serve as significant mediators in the relationship between ChatGPT and intention to purchase. These mediating variables reveal that the effect of ChatGPT is not purely a result of the content it delivers, but also how that content is perceived, processed, and emotionally interpreted by the tourist. The persistence of a significant direct effect alongside the indirect effects confirms that these are partial mediations, meaning that ChatGPT influences tourist behavior through both cognitive-emotional pathways and independent informational strength.

Theoretical implications

This study offers several theoretical contributions to the literature on AI-mediated consumer behavior in tourism by broadening the explanatory application of the S-O-R framework within a conversational AI context. Traditionally, the S-O-R model conceptualizes stimuli as environmental inputs that influence internal psychological states and subsequent behavioral responses. In the present study, conversational AI is positioned as an interactive technological stimulus capable of generating emotionally, cognitively, and contextually meaningful user experiences. The findings confirm that ChatGPT significantly influences tourist purchase intention, thereby supporting the fundamental S-O-R structure (Stimulus → Organism → Response). Rather than fundamentally redefining the S-O-R framework, this study enriches its explanatory relevance within AI-mediated tourism environments by identifying the complementary organism-level mechanisms through which conversational AI shapes behavioral intention. By incorporating perceived empathy, trust in AI-generated content, and content adaptability, the study provides a more comprehensive interpretation of the organism component through the simultaneous integration of emotional, cognitive, and contextual responses associated with AI interaction. This integrated perspective contributes to the literature by demonstrating that AI-mediated tourist behavior cannot be sufficiently explained through isolated psychological responses alone, but instead emerges through the interrelated operation of multiple complementary mechanisms within the organism stage.

The identification of perceived empathy as a significant outcome of AI interaction provides an important theoretical insight. The findings suggest that users may experience perceived emotional resonance even during interactions with non-human technological agents, particularly within conversational AI environments designed to simulate responsiveness and personalization. This observation contributes to expanding current discussions surrounding emotional processing within technology-mediated interactions and highlights the relevance of perceived empathy within digitally assisted tourism decision-making contexts. Accordingly, the study supports the growing view that emotional perceptions within AI interactions deserve greater consideration within behavioral and tourism-related models grounded in the S-O-R tradition.

Furthermore, the confirmed mediating roles of trust and content adaptability contribute to the theoretical understanding of how users interpret and evaluate AI-generated communication. These constructs connect technological capabilities with psychological interpretation, emphasizing that behavioral responses are shaped not only by informational quality but also by perceived reliability and contextual responsiveness. The simultaneous inclusion of emotional (perceived empathy), cognitive (trust in AI-generated content), and contextual (content adaptability) mechanisms provides a more integrated representation of the organism stage than is commonly addressed in prior tourism AI studies. Rather than examining these mechanisms independently, the present study demonstrates how they operate collectively in explaining tourist behavioral intention within conversational AI environments. In this respect, the study contributes a more conceptually integrated perspective for understanding AI-mediated tourist decision processes while remaining theoretically consistent with the core assumptions of the S-O-R framework.

In addition, this study contributes to bridging AI interaction research and tourism behavior literature by providing an empirically supported explanation of how AI-driven interactions translate into behavioral intentions. Rather than focusing exclusively on system performance or technological functionality, the findings emphasize the central role of user perception in shaping behavioral outcomes. This perspective reinforces the importance of psychological interpretation in explaining the effectiveness of conversational AI within tourism decision environments.

Finally, the identification of partial mediation effects provides additional insight into the complexity of AI-mediated behavioral responses. The findings indicate that conversational AI influences tourist purchase intention both directly and indirectly through internal psychological mechanisms. This dual influence suggests that AI-based stimuli may simultaneously trigger internal organismic responses while also exerting an independent persuasive effect on behavioral intention. Consequently, the study offers a more nuanced understanding of behavioral formation in technology-intensive tourism environments and further demonstrates the adaptability of the S-O-R framework for examining emerging forms of human–AI interaction within contemporary tourism research.

Practical implications

The findings of this study carry several practical implications for stakeholders in the tourism and hospitality industry as well as developers of AI-driven conversational systems. As tourists increasingly rely on digital technologies to inform their decisions, understanding how these tools shape behavior becomes essential for designing more effective, persuasive, and interactions perceived as emotionally resonant. For tourism organizations, the demonstrated direct and indirect influence of ChatGPT on tourists’ intention to purchase signals a strategic opportunity to integrate conversational AI into customer engagement touchpoints. By embedding tools like ChatGPT into travel websites, booking platforms, and mobile applications, service providers can offer tourists a more personalized and interactive planning experience—one that guides them seamlessly from information-seeking to purchasing. Moreover, leveraging AI to provide real-time, adaptive communication can reduce service load, enhance perceived service quality, and foster stronger digital engagement with prospective travelers. The findings further suggest that the combined integration of empathy-related interaction cues, trustworthy AI-generated content, and adaptive communication strategies can create a more comprehensive and effective digital service experience for tourists.

The study also underscores the importance of designing AI systems that not only deliver accurate content but also foster perceptions of simulated empathy, trust, and adaptability. For developers, this means moving beyond purely functional algorithms and focusing on the perceived emotional and psychological dimensions of interface design. For instance, programming ChatGPT to detect indicators of tourist sentiment, generate responses perceived as empathetic to user concerns, and adjust recommendations based on nuanced preferences can enhance both user satisfaction and conversion rates. AI systems designed to simulate emotionally responsive interaction does not require true consciousness but should simulate attentive and understanding communication patterns to strengthen the user experience. From a customer experience perspective, tourism providers should work with AI vendors to train models on domain-specific language, cultural nuances, and tourism-specific scenarios. This helps ensure that the AI is not only linguistically relevant but also perceived as contextually sensitive—two key components of perceived trust and adaptability. Moreover, transparency about how the AI works, how it generates responses, and how tourists can benefit from it will also help build trust and reduce skepticism regarding machine-generated content. Accordingly, tourism organizations should approach conversational AI implementation through an integrated strategy that simultaneously supports emotional reassurance, informational credibility, and contextual personalization rather than treating these dimensions separately.

Furthermore, the results suggest that a one-size-fits-all content strategy is no longer sufficient. Tourists expect AI systems to adapt to their unique needs, travel histories, and aspirations. Service providers should thus focus on data-driven personalization and real-time adaptation when deploying conversational AI in tourism environments. Features such as adjustable tone, multilingual capabilities, contextual memory, and feedback-driven learning can help AI systems maintain relevance across diverse customer segments and touchpoints. Finally, for tourism marketers and strategists, this study provides empirical evidence that AI tools are not merely cost-saving innovations, but revenue-generating assets. By influencing the tourist at both rational and emotional levels, ChatGPT and similar platforms can serve as AI-enabled digital sales support tools—driving higher engagement, stronger trust, and ultimately, higher conversion rates. Investment in AI systems capable of simulating emotionally responsive and context-adaptive interactions can thus deliver significant return on investment when deployed strategically.

Limitations and future research

While this study provides valuable insights into the psychological mechanisms through which ChatGPT influences tourist purchasing behavior, several methodological limitations should be acknowledged. First, the study adopted a cross-sectional design, capturing data at a single point in time. Although suitable for examining relationships between constructs, this approach limits the ability to establish causal inferences or assess how perceptions of ChatGPT evolve over time. Longitudinal research designs could offer a more dynamic perspective by tracking tourists’ interactions with conversational AI across different stages of the travel journey, from pre-trip planning to post-trip evaluation. Such an approach would enable a deeper understanding of how perceptions of trust, empathy, and adaptability develop with continued exposure.

Second, the study relied on a non-probability convenience sampling strategy targeting tourists with prior experience using ChatGPT for travel-related purposes. Although appropriate for the study’s objectives, this approach may limit the generalizability of the findings. Expanding future samples to include broader and more diverse populations—potentially through stratified or probabilistic sampling—would enhance external validity. Including both experienced and novice users could also reveal important moderating factors, such as technology readiness and digital literacy.

Third, the study focused exclusively on ChatGPT as the AI platform. While this focus ensures contextual relevance, it does not account for potential variations across different conversational AI systems. Comparative research examining multiple AI tools—such as Google Assistant, Trip.com chatbots, or proprietary hospitality bots—could provide deeper insights into how system-specific characteristics influence perceived empathy, trust, and adaptability.

Fourth, although demographic variables were collected, their potential moderating effects were not examined. This limits the ability to understand how different user segments may respond to AI-mediated interactions. Future research could employ multi-group analysis to explore variations across age, gender, cultural background, and levels of technological familiarity, thereby offering more nuanced and actionable insights.

Finally, the study focused on three key mediating variables—perceived empathy, trust in AI-generated content, and content adaptability. While these constructs capture essential dimensions of AI interaction, other psychological and contextual factors may also play a role. Variables such as perceived enjoyment, anthropomorphism, perceived control, and situational involvement may further enrich the explanatory power of the model. Expanding the organism component of the S-O-R framework to incorporate these dimensions would support the development of more comprehensive models of AI-driven tourist behavior.

Conclusion

The rapid integration of conversational AI into tourism and hospitality services is reshaping how tourists search for information, evaluate alternatives, and make purchase decisions. Within this evolving landscape, the present study demonstrated that ChatGPT functions as more than a technological information source; it represents an influential interactive stimulus capable of shaping tourist behavior through multiple psychological pathways. By applying the Stimulus–Organism–Response (S-O-R) framework, the study revealed that tourists’ behavioral intentions are significantly influenced by how AI-generated interactions are emotionally interpreted, cognitively trusted, and contextually adapted to individual needs. The findings further highlight that perceived empathy, trust in AI-generated content, and content adaptability operate collectively as complementary mechanisms that translate conversational AI interaction into purchase-related outcomes. In doing so, the study contributes to advancing the theoretical understanding of AI-mediated tourist behavior while emphasizing the growing strategic importance of conversational AI in tourism service ecosystems. As AI technologies continue to evolve, the ability of tourism organizations to design interactions that are trustworthy, adaptive, and psychologically engaging will likely become a critical determinant of competitive advantage and customer conversion in digital tourism environments.

Footnotes

ORCID iDs

Ahmed Mohamed Hasanein

Bassam Samir Al-Romeedy

Ethical considerations

This research complies with all relevant guidelines and regulations for studies involving human subjects.

Author contributions

Bassam Samir Al-Romeedy: Conceptualization, study design, and data collection. Ahmed Mohamed Hasanein: Data analysis, interpretation of results, and visualization. Aya Ahmed Abdel Majeed: Literature review, theoretical framework development, and drafting the manuscript. Abdullah H. Seraj: Supervision, critical revisions, and final editing of the manuscript. All authors contributed to the research and approved the final version of the paper.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the Deanship of Scientific Research, Vice Presidency for Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia, grant number [KFU261792].

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Morrison

. Marketing and managing tourism destinations. Routledge, 2023.

Al-Romeedy

Singh

. From search to stay: mapping the digital journey of the modern traveler. In: Azman

Valeri

Albattat

, et al. (eds). Decoding tourist behavior in the digital era: insights for effective marketing. IGI Global, 2025, pp. 283–308. https://doi.org/10.4018/979-8-3693-3972-5.ch012

Diaz

Carranza

Martin-Consuegra

, et al. Technology-driven smart consumer experiences in tourism destinations. Curr Issues Tourism 2024; 29: 967–997. https://doi.org/10.1080/13683500.2024.2443039

Bansal

Chamola

Hussain

, et al. Transforming conversations with AI—A comprehensive study of ChatGPT. Cogn Comput 2024; 16(5): 2487–2510.

Stergiou

Nella

. ChatGPT and tourist decision‐making: an accessibility–diagnosticity theory perspective. Int J Tourism Res 2024; 26(5): e2757.

Sridhar

. ChatGPT and artificial intelligenÎ. Academic Guru Publishing House, 2024.

Sigala

Ooi

Tan

, et al. Understanding the impact of ChatGPT on tourism and hospitality: trends, prospects and research agenda. J Hospit Tourism Manag 2024; 60: 384–390.

Batouei

Nikbin

Foroughi

. Acceptance of ChatGPT as an auxiliary tool enhancing travel experience. J Hospit Tour0020Insights 2025; 8: 2744–2763. https://doi.org/10.1108/JHTI-10-2024-1142

Farahat

. Applications of artificial intelligence as a marketing tool and their impact on the competitiveness of the Egyptian tourist destination (doctoral dissertation). Minia University, 2023.

10.

Foroughi

Iranmanesh

Yadegaridehkordi

, et al. Factors affecting the use of ChatGPT for obtaining shopping information. Int J Consum Stud 2025; 49(1): e70008.

11.

Shalan

. The impact of artificial intelligence on improving tourism service quality in the Egyptian destination. Doctoral Dissertation, University of Sadat City, Egypt, 2025.

12.

Shi

Lin

Wang

, et al. Sociable robots or focused speakers? Transforming customer experience with communication style and embodiment type in smart home devices adoption. Behav Inf Technol 2025; 44: 3559–3574. https://doi.org/10.1080/0144929X.2025.2485401

13.

Cambra-Fierro

Chabán

Fuentes-Blasco

, et al. Exploring faculty adoption of ChatGPT: a stimuli-organism-response approach. Interact Learn Environ 2025; 33: 5580–5598. https://doi.org/10.1080/10494820.2025.2483415

14.

Zaki

Al-Romeedy

. Chatbot symbolic recovery and customer forgiveness: a moderated mediation model. J Hosp Tour Technol 2024; 15(4): 610–628.

15.

Leveau

. Tourism, technologies and consumption in the 5.0 era: discovery and reflection in the face of organizational and societal issues. John Wiley & Sons, 2025.

16.

Han

Kim

Hailu

, et al. Effects of cognitive, affective and normative drivers of artificial intelligence ChatGPT on continuous use intention. J Hosp Tour Technol 2024; 15(4): 629–647.

17.

Pham

Duong

Nguyen

. What drives tourists’ continuance intention to use ChatGPT for travel services? A stimulus002Dorganism-response perspective. J Retailing Consum Serv 2024; 78: 103758.

18.

Pillai

Sivathanu

Dwivedi

. Shopping intention at AI-powered automated retail stores (AIPARS). J Retailing Consum Serv 2020; 57: 102207.

19.

Tsumura

Yamada

. The role of individual recognition in shaping empathy and trust toward an agent. PLoS One 2025; 20(7): e0327329.

20.

Chen

Wei

. Role of artificial intelligence in travel decision making and tourism product selling. Asia Pac J Tourism Res 2024; 29(3): 239–253.

21.

Shin

Kim

Lee

, et al. ChatGPT for trip planning: the effect of narrowing down options. J Trav Res 2025; 64(2): 247–266.

22.

Martínez

Herrero

García

. Communicating destination social responsibility through social media: the roles of tourists’ social engagement, citizenship behaviors, and emotions. J Trav Res 2025; 64(4): 929–949.

23.

Law

Lovett

, et al. Tourist acceptance of ChatGPT in travel services: the mediating role of parasocial interaction. J Trav Tourism Market 2024; 41(7): 955–972.

24.

Luo

Hasan

Zamri bin Ahmad

, et al. Influence of short video content on consumers purchase intentions on social media platforms with trust as a mediator. Sci Rep 2025; 15(1): 16605.

25.

Shalan

Gaafar

Al-Romeedy

, et al. The impact of artificial intelligence on improving tourism service quality in the Egyptian destination. J. Fac. Tour. Hotel Univ. Sadat City 2024; 8(2): 1–26.

26.

Yang

Zhang

Feng

. Personalized tourism recommendations and the e-tourism user experience. J Trav Res 2024; 63(5): 1183–1200.

27.

Dixit

. Marketing in hospitality and travel. Educohack Press, 2025.

28.

Sultan

Hussain

. Towards a sustainable workforce: integrating workplace spirituality, green leadership, and employee adaptability for green creativity. Discover Sustainability 2025; 6(1): 318.

29.

Lee

Shin

, et al. Revisiting and exploring trust in the digital era: conceptualization and scale development of digital trust in hospitality and tourism. Inf Technol Tourism 2025; 27(1): 125–155.

30.

Patil

Rane

. The future impact of ChatGPT on several business sectors. Deep Science Publishing, 2024.

31.

Çolak

. The role of generative pre-trained transformers (GPT) in recreational tourism: an interview with ChatGPT. Journal of Sport Sciences Research 2023; 8(3): 733–748.

32.

Rajesh

. ChatGPT transforming industries through AI. Academic Guru Publishing House, 2024.

33.

Al-Romeedy

Singh

. Reimagining customer relations through AI innovation. In: Darwish

Kumar

(eds). Utilizing AI and Machine Learning in Financial Analysis. IGI Global, 2025, pp. 351–368. https://doi.org/10.4018/979-8-3693-8507-4.ch019

34.

Wang

Miao

Xiong

, et al. How does ChatGPT-generated information influence travel planning? The mediating role of inspiration. Tour Rev 2025; 81: 694–717. https://doi.org/10.1108/TR-03-2025-0235

35.

Correia

Martins

Fontes

. AI innovations for travel and tourism. IGI Global, 2024.

36.

Ferrer-Rosell

Massimo

Berezina

. Information and communication technologies in tourism 2023: proceedings of the ENTER 2023 eTourism conference, january 18-20, 2023. Springer Nature, 2023. https://doi.org/10.1007/978-3-031-25752-0

37.

Zhu

Song

Duan

. Emotional arousal: how artificial intelligence-generated content influences tourism decision-making. Tour Recreat Res 2024; 50: 1639–1650. https://doi.org/10.1080/02508281.2024.2439321

38.

Xinlin

Wenting

Shah

, et al. Transforming hospitality decision-making: the impact of generative AI on cognitive alignment and adaptive intelligence. J Qual Assur Hospit Tourism. 2025; 1–35. https://doi.org/10.1080/1528008X.2025.2528044

39.

Yankouskaya

Liebherr

Ali

. Can ChatGPT be addictive? A call to examine the shift from support to dependence in AI conversational large language models. Human-Centric Intelligent Systems 2025; 5: 77–89.

40.

Maghrifani

Sneddon

Liu

. Personal values and travel motivations: the moderating effects of visit experience, gender and age. J Hospit Tour Insights 2024; 7(5): 3091–3109.

41.

Chakraborty

Biswal

. Is ChatGPT a responsible communication: a study on the credibility and adoption of conversational artificial intelligence. J Promot Manag 2024; 30(6): 929–958.

42.

Shin

. Algorithms, humans, and interactions: how do algorithms interact with people? Designing meaningful AI experiences. Taylor & Francis, 2023.

43.

Virvou

. Artificial intelligence and user experience in reciprocity: contributions and state of the art. Intell Decis Technol 2023; 17(1): 73–125.

44.

Rostami

Navabinejad

. Artificial empathy: user experiences with emotionally intelligent chatbots. AI and Tech in Behavioral and Social Sciences 2023; 1: 19–27.

45.

Damiasih

. Psychological factors influencing tourist decision making: a qualitative review in the context of business psychology. Golden Ratio of Social Science and Education 2025; 5(1): 59–72.

46.

Gude

. Factors influencing ChatGPT adoption for product research and information retrieval. J Comput Inf Syst 2025; 65(2): 222–231.

47.

Vereschak

Bailly

Caramiaux

. How to evaluate trust in AI-assisted decision making? A survey of empirical methodologies. Proc ACM Hum-Comput Interact 2021; 5(CSCW2): 1–39.

48.

Park

Son

. AI recommendation vs. crowdsourced recommendation vs. travel expert recommendation: the moderating role of consumption goal on travel destination decision. PLoS One 2025; 20(3): e0318719.

49.

Chiu

. Exploring user awareness and perceived usefulness of generative AI in higher education: the moderating role of trust. Educ Inf Technol 2025; 30: 21317–21351. https://doi.org/10.1007/s10639-025-13612-7

50.

Alam

Ahsan

Kokash

, et al. Live streaming characteristics, motivational factors, trust and impulsive hotel booking intention: mediating role of customer engagement and moderating role of time pressure. J Hospit Market Manag 2025; 34(5): 651–680.

51.

Gao

Liu

. The effect of airline ticket blind box design on consumer’s impulsive purchase intention: the perspective of expectation confirmation theory. Curr Issues Tourism 2025; 1: 1–11. https://doi.org/10.1080/13683500.2025.2462159

52.

Del Vecchio

Lazoi

Marche

, et al. Designing innovative digital solutions in the cultural heritage and tourism industry: best practices for an immersive user experience. Appl Sci 2025; 15(9): 4935.

53.

Jiang

Wan

, et al. User-centric design experience or privacy risk perception? The dual pathways of tourism platform’s AI capability to perceived user value. Curr Issues Tourism 2024; 29: 1376–1393. https://doi.org/10.1080/13683500.2024.2394569

54.

Shalan

Gaafar

Al-Romeedy

, et al. The role of artificial intelligence in tourism service quality: insights from employees. J. Fac. Tour. Hosp 2024; 8(2/2): 102–119.

55.

Paul

Ueno

Dennis

. ChatGPT and consumers: benefits, pitfalls and future research agenda. Int J Consum Stud 2023; 47(4): 1213–1225.

56.

Agarwal

Mitra

. What drives users impulse buying behavior in social media advertising? The roles of empathy expression and privacy concerns. J Market Theor Pract 2025; 34: 479–500. https://doi.org/10.1080/10696679.2025.2452848

57.

Abou Hashish

. Compassion through technology: Digital empathy concept analysis and implications in nursing. Digit Health 2025; 11: 20552076251326221.

58.

Abou-Shouk

Abdelhakim

Elgarhy

, et al. ChatGPT usage intention for tourism and hospitality customers. Tour Recreat Res 2025; 51: 533–550. https://doi.org/10.1080/02508281.2025.2465962

59.

Zhang

Liang

. Understanding virtual agents’ service quality in the context of customer service: a fit-viability perspective. Electron Commer Res Appl 2024; 65: 101380.

60.

Hair

Hult

Ringle

, et al. A primer on partial least squares structural equation modeling (PLS-SEM). Sage publications, 2021.

61.

Gao

Zhang

. Matching emotion regulation strategies with specific emotions in tourist experiences. J Hospit Tour Insights 2025; 8(2): 553–571.

62.

Kodal

. The role of leadership and emotional intelligence in destination management: a conceptual review. Tourist Destination 2024; 2(1): 1–12.

63.

Sousa

Cardoso

Dias

. The use of artificial intelligence systems in tourism and hospitality: the tourists’ perspective. Adm Sci 2024; 14(8): 165.

64.

Ferreira

Pereira

. More than brands, friends: enhancing emotional engagement through AI in digital relationship marketing context. In: Leveraging AI for effective digital relationship marketing. IGI Global‏, 2025, pp. 287–308.

65.

Huang

Tian

Zhang

, et al. Parallel tourism system with foundation models and decentralized autonomous organizations. Int J Tourism Res 2025; 27(2): e70008.

66.

Lee

. Navigating the generative AI travel landscape: the influence of ChatGPT on the evolution from new users to loyal adopters. Int J Contemp Hospit Manag 2025; 37(4): 1421–1447.

67.

Balaskas

Tsiantos

Chatzifotiou

, et al. Determinants of ChatGPT adoption intention in higher education: expanding on TAM with the mediating roles of trust and risk. Information 2025; 16(2): 82.

68.

Ramos

. User-generated content and its impact on purchase intent for tourism products: a comparative analysis of millennials and centennials on TikTok. Future Internet 2025; 17(3): 105.

69.

Ilieva

Yankova

Klisarova-Belcheva

. Effects of generative AI in tourism industry. Information 2024; 15(11): 671.

70.

Singh

Wongmahesak

Chandra

. Content marketing with search engine optimization: fostering successful businesses in online marketing 6.0. PSCSOM 2025; 6: 17–36.