Artificial Intelligence in Travel Planning: Usage Behavior, Perceived Value, and Continuance Intention Among Korean Travelers

1. Introduction

The adoption of artificial intelligence (AI) in everyday consumer decision-making has accelerated rapidly since the public release of large language model applications such as ChatGPT. The development of generative AI has expanded the role of AI from traditional information retrieval and automated recommendation systems to interactive tools capable of understanding natural-language questions, generating personalized responses, and assisting users throughout the decision-making process. In the travel sector specifically, AI-powered tools now assist travelers with destination discovery, itinerary construction, accommodation and flight search, transportation information, and personalized recommendations, functioning in effect as a conversational travel assistant available at different stages of the travel experience. Rather than relying solely on conventional search engines or travel websites, travelers can now interact directly with AI systems to compare destinations, organize travel schedules, identify places of interest, and obtain suggestions based on their individual preferences (Bo & Tiesen, 2026).

This transformation is particularly relevant to travel planning because tourism decisions involve large amounts of information and multiple interconnected choices. Travelers must often evaluate destinations, transportation options, accommodation, restaurants, attractions, costs, schedules, and available activities before making a final decision. AI tools can potentially reduce the time and effort required to process this information by integrating multiple types of travel-related information into a single conversational interface. Industry reports indicate that a majority of travelers worldwide already rely on AI-powered platforms during the recommendation and planning phase of a trip, even as concerns about the accuracy, reliability, and trustworthiness of AI-generated information persist. These concerns are important because inaccurate recommendations, outdated information, or fabricated details can directly affect travelers’ financial decisions and overall travel experiences (Hakseung et al., 2026).

The Technology Acceptance Model (TAM) proposed by Davis (1989) remains one of the most widely applied frameworks for explaining why individuals adopt a given technology, positing that perceived usefulness and the resulting attitude toward a system are key determinants of technology acceptance and continued use. From this perspective, travelers are more likely to adopt AI-based travel tools when they believe that such tools provide meaningful benefits and improve the efficiency of their travel-planning activities. However, usefulness alone may not fully explain continued adoption. Travelers must also perceive the information provided by AI as sufficiently reliable, convenient, personalized, and trustworthy to incorporate it into future decision-making (Halibas et al., 2026).

Building on this logic, recent tourism research has shown that trust in AI-generated recommendations develops through a heuristic–systematic evaluation process and is a significant predictor of travelers’ intention to rely on AI-based recommendation systems (Shi, Gong, & Gursoy, 2021). At the same time, studies on AI and robotics adoption in hospitality and tourism caution that resistance to AI persists where users doubt the accuracy, transparency, or personalization of the information provided (Goel, Kaushik, Sivathanu, Pillai, & Vikas, 2022). These findings suggest that the relationship between actual AI usage and future adoption is not determined solely by technological availability. Instead, users’ experiences and perceptions of the value delivered by AI may play an important role in determining whether they continue to use these tools (Hong et al., 2024).

The increasing use of AI in tourism also raises questions about differences among travelers. Individual characteristics such as age, occupation, travel frequency, and previous experience with AI may influence how travelers interact with AI-based travel services and how much value they perceive from them. Travelers who frequently travel overseas may have greater experience with online travel information and therefore use AI for a wider range of planning tasks, while frequent AI users may be more familiar with the capabilities and limitations of generative AI. Examining these differences can provide a more detailed understanding of how AI adoption varies across traveler groups rather than treating travelers as a single homogeneous population (Hongyoul, 2026).

South Korea is one of the most digitally connected travel markets in the world, with high smartphone penetration and a population that readily experiments with new digital services. Korean travelers increasingly have access to AI-based tools through smartphones, search platforms, travel applications, and generative AI services. This environment provides a relevant context for examining how AI is incorporated into actual travel-planning behavior. Understanding how Korean travelers use AI across different trip-planning tasks, how favorably they perceive its value, and whether these factors translate into an intention to keep using AI in the future carries direct relevance for destination marketers, travel technology providers, hospitality businesses, and tourism policy makers seeking to design AI-enabled services that travelers will trust and continue to use (Arpaci et al., 2026).

Therefore, this study focuses on the relationship among AI usage behavior, perceived value, and continuance intention among Korean travelers with overseas travel experience. By examining AI use across several practical travel-planning activities and comparing perceptions across respondent characteristics, the study seeks to provide empirical evidence regarding the role of AI in contemporary travel decision-making. The findings may contribute to the growing tourism literature on generative AI adoption while also providing practical implications for the development of more personalized, convenient, credible, and user-oriented AI travel services.

1.1. Objectives of the Study

This study aimed to (1) describe the demographic and travel profile of respondents in terms of age, gender, occupation, and recent overseas travel frequency, as well as their pattern of AI use before and during trips; (2) determine the extent to which respondents use AI across five trip-planning tasks — destination search, itinerary planning, accommodation search, transportation search, and activity/restaurant recommendation; (3) determine respondents’ perceived value of AI in terms of time-saving, personalization, convenience, and trust; (4) test whether AI usage behavior, perceived value, and continuance intention differ significantly when respondents are grouped according to their demographic, travel, and AI-use profiles; and (5) test the relationship between AI usage behavior, perceived value, and travelers’ continuance intention to use AI in future trip planning (Im & Ha, 2026).

More specifically, the study seeks to determine whether travelers’ actual interaction with AI across different planning activities is associated with their evaluation of AI’s practical value and their willingness to continue using AI-based travel tools. By examining these relationships, the study aims to provide a clearer empirical understanding of the factors associated with continued AI use among Korean travelers and to identify areas that may be relevant to the future development of AI-supported travel services (Kim et al., 2024).

2. Methodology

2.1. Research Design

This study used a descriptive-correlational research design to examine the use and continued adoption of artificial intelligence (AI) among Korean travelers. The descriptive component profiled respondents’ demographic and travel characteristics and described the extent of their AI usage behavior and perceived value of AI in travel planning. This component provided an overview of how respondents interact with AI across different travel-related activities and how they evaluate the practical benefits of AI-based tools (Choe et al., 2024).

The correlational component examined the relationships among AI usage behavior, perceived value, and continuance intention to use AI for future travel planning. The study did not manipulate any variables or introduce an experimental intervention; instead, it examined naturally occurring patterns based on respondents’ reported experiences and perceptions. This design was considered appropriate because the study sought to describe existing AI usage patterns and determine whether meaningful statistical associations exist among the major study variables (Davis, 1989).

2.2. Respondents of the Study

The respondents were 401 Korean adults who had prior overseas travel experience and were reached through a structured online questionnaire. Respondents were included because previous overseas travel experience provided a relevant basis for evaluating AI use in actual travel planning rather than relying exclusively on hypothetical situations. The online survey format also allowed respondents with different demographic and travel backgrounds to participate and report their recent experiences with AI-supported travel information and planning (Jung-eun, 2026).

The great majority of respondents (98.8%) reported at least one overseas trip within the past three years, while 95.3% reported having used AI for a travel-related purpose at least once. These figures indicate that the sample was well positioned to provide information about actual AI usage experience rather than hypothetical intention alone. The respondents’ recent travel experience and relatively high level of AI exposure also provided an appropriate basis for examining differences in AI usage behavior and perceived value across demographic and travel-related groups (Linnes et al., 2026).

2.3. Instrument of the Study

A structured questionnaire was used as the primary data-collection instrument. The questionnaire was organized into three parts to correspond with the major variables and objectives of the study. Part I gathered demographic and travel-related profile information, including characteristics such as age, gender, occupation, overseas travel frequency, and AI-use frequency. These variables were subsequently used to examine whether patterns of AI usage, perceived value, and continuance intention varied across respondent groups (Tuyen, 2026).

Part II measured AI usage behavior across five specific travel-planning tasks: destination search, itinerary planning, accommodation search, transportation search, and activity/restaurant recommendation. These items were designed to capture the extent to which respondents incorporate AI into practical travel-planning activities rather than measuring general awareness of AI alone. Part III measured the perceived value of AI in terms of time-saving, personalization, convenience, and trust, together with travelers’ continuance intention to use AI in future travel planning (Xu et al., 2025).

Items in Parts II and III used a 5-point Likert scale ranging from “Strongly Disagree” (1) to “Strongly Agree” (5), allowing respondents to indicate the degree to which each statement reflected their experience or perception. Composite scores were subsequently calculated for the major constructs to facilitate statistical analysis. Internal consistency was excellent for both the AI usage behavior scale (Cronbach’s α = .964, 5 items) and the perceived value scale (Cronbach’s α = .945, 4 items). Both coefficients were substantially above the generally accepted threshold of .70, indicating a high level of internal consistency among the items measuring each construct (Wong et al., 2025).

2.4. Data Analysis

The collected data were analyzed using descriptive and non-parametric statistical procedures consistent with the objectives and distributional characteristics of the study variables. Frequency and percentage distributions were used to describe the demographic, travel, and AI-use profiles of the respondents. These descriptive statistics provided an overview of the composition of the sample and the respondents’ recent travel and AI-use experiences (Tay et al., 2026).

Weighted means and ranking were used to describe the level of AI usage behavior and perceived value across the measured dimensions. These analyses helped identify which travel-planning activities were most commonly associated with AI use and which perceived benefits of AI were evaluated more favorably by respondents (Silalahi et al., 2025).

Before conducting inferential analyses, the normality of the composite scores was examined using the Shapiro-Wilk test. The results indicated that the composite scores for AI usage behavior, perceived value, and continuance intention were not normally distributed (p < .001 for all variables). Therefore, non-parametric statistical procedures were selected because they do not require the assumption of normally distributed data. The Mann-Whitney U test was used to compare two-group variables, specifically gender, while the Kruskal-Wallis H test was used for variables with three or more groups, including age, occupation, travel frequency, and AI-use frequency (Teng et al., 2026).

Finally, Spearman rank-order correlation (rho) was employed to examine the direction and strength of the relationships among AI usage behavior, perceived value, and continuance intention. This analysis was appropriate for assessing associations among the study variables given the non-normal distribution of the composite scores and the ordinal nature of the underlying Likert-scale responses. An alpha level of .05 was used for all inferential tests. Statistical significance was therefore determined based on p-values below .05, while the correlation coefficients were interpreted in terms of the strength and direction of the observed relationships (Siamak et al., 2025).

3. Results and Discussion

3.1. Demographic and Travel Profile

Table 1 presents the profile of the 401 respondents.

Most respondents were between 45 and 54 years old (41.9%), followed by other age groups, indicating that the sample included a substantial proportion of relatively experienced adult travelers. Male respondents accounted for 57.0% of the sample. In terms of occupation, office workers represented the largest group (26.9%), followed by self-employed respondents (23.9%), suggesting that the sample consisted largely of economically active adults with relatively regular engagement in work and travel-related decision-making. Regarding overseas travel frequency, 69.9% of respondents had traveled abroad between 3 and 10 times during the past three years, indicating that a considerable proportion of the sample had repeated international travel experience rather than being occasional first-time travelers.

AI use in travel planning was widespread: 95.3% of respondents had used AI for a travel-related purpose, demonstrating a high level of practical exposure to AI-based travel services within the sample. Approximately half of the respondents (50.4%) primarily used ChatGPT, while Google Gemini was the second most commonly used platform (37.9%). The relatively high use of both platforms suggests that generative AI applications have become accessible and relevant to travelers when searching for information and preparing for overseas trips. Respondents’ repeated travel experience, combined with their relatively high level of AI use, provides an appropriate context for examining actual AI usage behavior and perceived value rather than relying solely on hypothetical expectations (Arora et al., 2025).

This pattern is consistent with the growing body of research on AI adoption in tourism, which indicates that generative AI is increasingly being incorporated into travel information search, recommendation, and planning activities (Goel et al., 2022). The findings further suggest that, within this sample, AI has moved beyond being perceived merely as a novel technological application and has become a practical travel-planning tool used alongside conventional sources of travel information. The high proportion of respondents with previous AI experience also provides a useful basis for examining whether repeated interaction with AI is associated with more favorable perceptions and stronger intentions to continue using AI for future travel planning.

3.2. AI Usage Behavior in Travel Planning

Respondents reported that they “agree” that they use AI across all five major trip-planning tasks, with a composite mean of 3.79, indicating an overall favorable level of AI usage in travel planning. AI was used most frequently for obtaining restaurant and attraction recommendations (M = 3.81), closely followed by hotel search and transportation search (M = 3.79 each). These results indicate that AI is being incorporated into a variety of practical travel-planning activities rather than being limited to a single stage of the decision-making process. In particular, the relatively high use of AI for restaurant and attraction recommendations suggests that travelers value AI’s ability to generate multiple options and tailor suggestions according to their interests, location, and travel preferences.

This pattern further suggests that Korean travelers rely on AI mainly during the discovery and comparison stages of planning, where a large volume of potentially relevant options must be identified, organized, and filtered efficiently. AI can reduce the time required to search through numerous websites or platforms by presenting recommendations in a conversational and easily accessible format. This finding is broadly consistent with research indicating that AI is particularly useful for discovery-type travel tasks, such as identifying activities, attractions, restaurants, and other destination-related options, rather than serving as the sole basis for final validation of a decision.

Itinerary planning, which requires integrating multiple constraints such as available time, budget, transportation routes, opening hours, and the sequencing of activities, was used slightly less frequently (M = 3.77). Although the difference was relatively small, this result may reflect the more complex and judgment-intensive nature of itinerary construction. Travelers may use AI to generate an initial itinerary or provide alternative arrangements while still relying on their own judgment to determine whether the proposed schedule is realistic and suitable. Overall, the relatively narrow range of mean scores across all five tasks indicates that AI has become a broadly applicable planning resource, with respondents reporting relatively consistent use across destination-related search, accommodation, transportation, and activity planning (Dhiman & Jamwal, 2023).

Table 2. AI usage behavior across travel-planning tasks. 

Task	WM	VI	Rank
I use AI to receive recommendations for restaurants or tourist attractions.	3.81	Agree	1
I use AI to search for hotels or accommodations.	3.79	Agree	2.5
I use AI to check flight or transportation information.	3.79	Agree	2.5
I use AI to search for information about travel destinations.	3.78	Agree	4
I use AI to plan a travel itinerary.	3.77	Agree	5

Legend: 4.20–5.00 = Strongly Agree; 3.40–4.19 = Agree; 2.60–3.39 = Neutral; 1.80–2.59 = Disagree; 1.00–1.79 = Strongly Disagree. Composite mean = 3.79 (Agree), SD = 0.93.

Table 2. AI usage behavior across travel-planning tasks. 

Task	WM	VI	Rank
I use AI to receive recommendations for restaurants or tourist attractions.	3.81	Agree	1
I use AI to search for hotels or accommodations.	3.79	Agree	2.5
I use AI to check flight or transportation information.	3.79	Agree	2.5
I use AI to search for information about travel destinations.	3.78	Agree	4
I use AI to plan a travel itinerary.	3.77	Agree	5

Legend: 4.20–5.00 = Strongly Agree; 3.40–4.19 = Agree; 2.60–3.39 = Neutral; 1.80–2.59 = Disagree; 1.00–1.79 = Strongly Disagree. Composite mean = 3.79 (Agree), SD = 0.93.

3.3. Perceived Value of AI in Travel Planning

The item rated highest was personalization, with respondents agreeing that AI provides information matching their personal preferences (M = 3.83). This was followed closely by time savings (M = 3.81), while convenience and trust received identical mean scores (M = 3.80 each). The narrow spread between these means, only 0.03 points, indicates that respondents perceived the major benefits of AI at a remarkably similar level. In other words, travelers did not view AI solely as a tool for obtaining information more quickly; they also recognized its ability to provide recommendations that are relevant to their individual interests and to simplify the overall travel-planning process. The high score for personalization is particularly notable because travel preferences can vary considerably in terms of destination interests, budget, schedule, food preferences, and preferred activities.

The results also suggest that the practical and relational dimensions of AI are closely connected. For this sample, the efficiency-related benefits of AI, such as speed and convenience, and the more relational benefit of trust were perceived almost equally strongly. Travelers therefore appear to evaluate AI not simply according to how quickly it produces information, but also according to whether the information is sufficiently relevant and credible to be useful for actual travel decisions. This pattern is consistent with prior research showing that trust in AI-based travel recommendations can develop through both systematic evaluation of information quality and a more heuristic, convenience-driven response to the technology (Shi et al., 2021). The findings consequently support the view that perceived value in AI-supported travel planning consists of multiple interconnected dimensions rather than a single perception of technological usefulness.

Notably, respondents also “agreed” that they were concerned that AI-provided travel information might be inaccurate (M = 3.78, SD = 1.02). This score was almost identical in magnitude to their agreement that they trusted AI, suggesting that trust and concern about accuracy can coexist within the same user. Rather than necessarily representing a contradiction, this pattern may reflect a form of “calibrated trust,” in which travelers recognize the usefulness of AI while remaining aware that its outputs may require verification. Such awareness is particularly relevant in travel planning because information concerning prices, opening hours, transportation schedules, reservations, and local conditions can change rapidly.

Furthermore, the concern about accuracy was positively and strongly correlated with both AI usage behavior (ρ = .707) and perceived value (ρ = .803). This indicates that respondents who reported greater engagement with AI and higher perceived value were also more likely to recognize its potential limitations. Greater familiarity with AI may therefore increase not only reliance on the technology but also users’ awareness of situations in which its information should be checked against other sources. Rather than necessarily reducing engagement, awareness of potential inaccuracies may encourage travelers to use AI as an initial planning and recommendation tool while retaining personal judgment and external verification for important travel decisions. This combination of high perceived value and continued awareness of limitations provides an important perspective on how travelers may develop sustainable patterns of AI use in travel planning (Shi et al., 2021).

Table 3. Perceived value of AI in travel planning. 

Statement	WM	VI	Rank
AI provides travel information that matches my personal preferences.	3.83	Agree	1
AI reduces the time required to plan a trip.	3.81	Agree	2
Using AI makes travel planning more convenient.	3.80	Agree	3.5
I tend to trust the travel information provided by AI.	3.80	Agree	3.5

Legend: 4.20–5.00 = Strongly Agree; 3.40–4.19 = Agree; 2.60–3.39 = Neutral; 1.80–2.59 = Disagree; 1.00–1.79 = Strongly Disagree. Composite mean = 3.81 (Agree), SD = 0.94.

Table 3. Perceived value of AI in travel planning. 

Statement	WM	VI	Rank
AI provides travel information that matches my personal preferences.	3.83	Agree	1
AI reduces the time required to plan a trip.	3.81	Agree	2
Using AI makes travel planning more convenient.	3.80	Agree	3.5
I tend to trust the travel information provided by AI.	3.80	Agree	3.5

Legend: 4.20–5.00 = Strongly Agree; 3.40–4.19 = Agree; 2.60–3.39 = Neutral; 1.80–2.59 = Disagree; 1.00–1.79 = Strongly Disagree. Composite mean = 3.81 (Agree), SD = 0.94.

3.4. Differences in AI Usage Behavior, Perceived Value, and Continuance Intention by Profile

AI usage behavior, perceived value, and continuance intention did not differ significantly by gender, suggesting that male and female Korean travelers in this sample reported broadly comparable patterns of AI adoption and future-use intention. In contrast, significant differences emerged across age, occupation, overseas travel frequency, and pre-trip AI use frequency. Respondents aged 55 and older reported the highest AI usage behavior and perceived value (M = 4.14 and 4.13, respectively), while the youngest group (18–24) reported the lowest levels (M = 3.15 and 3.26). This finding is noteworthy because it differs from the common assumption that younger, “digital native” travelers necessarily represent the strongest users of emerging AI technologies. In this sample, older travelers reported greater engagement with AI and more favorable evaluations of its value. One possible explanation is that experienced adult travelers may perceive AI as particularly useful for reducing the time and effort required to organize overseas travel, especially when planning involves unfamiliar destinations, multiple transportation options, or a large amount of travel information. However, the cross-sectional nature of the study does not establish that age itself causes higher AI usage.

Differences were also observed according to occupation. Professionals reported the highest level of AI usage and perceived value (M = 4.18), followed by self-employed respondents (M = 4.08), whereas university students reported the lowest level (M = 3.20). This pattern may reflect differences in travel experience, available time, financial resources, or the perceived opportunity cost of spending time on travel planning. Working professionals and self-employed travelers may place greater value on AI’s ability to quickly organize information and reduce the amount of time required for planning. University students, by comparison, may rely more heavily on alternative information sources such as social media, online communities, or recommendations from peers. These possible explanations should be interpreted cautiously because the present study measured associations between respondent characteristics and AI-related variables rather than the underlying reasons for those differences.

Table 4. Comparison of AI usage behavior, perceived value, and continuance intention when grouped by profile. 

Profile Variable	Variable	Statistic	p-value	Interpretation
Gender	AI Usage Behavior	U = 20040.0	.699	Not Significant
	Perceived Value	U = 19863.5	.820	Not Significant
	Continuance Intention	U = 19554.0	.961	Not Significant
Age	AI Usage Behavior	H = 28.02	<.001	Significant
	Perceived Value	H = 20.93	<.001	Significant
	Continuance Intention	H = 16.71	.002	Significant
Occupation	AI Usage Behavior	H = 36.83	<.001	Significant
	Perceived Value	H = 32.22	<.001	Significant
	Continuance Intention	H = 27.14	<.001	Significant
Overseas travel frequency	AI Usage Behavior	H = 139.77	<.001	Significant
	Perceived Value	H = 135.72	<.001	Significant
	Continuance Intention	H = 110.40	<.001	Significant
AI use frequency (pre-trip)	AI Usage Behavior	H = 191.58	<.001	Significant
	Perceived Value	H = 146.96	<.001	Significant
	Continuance Intention	H = 127.23	<.001	Significant

Legend: Significant at p-value < .05. U = Mann-Whitney U statistic; H = Kruskal-Wallis H statistic.

Table 4. Comparison of AI usage behavior, perceived value, and continuance intention when grouped by profile. 

Profile Variable	Variable	Statistic	p-value	Interpretation
Gender	AI Usage Behavior	U = 20040.0	.699	Not Significant
	Perceived Value	U = 19863.5	.820	Not Significant
	Continuance Intention	U = 19554.0	.961	Not Significant
Age	AI Usage Behavior	H = 28.02	<.001	Significant
	Perceived Value	H = 20.93	<.001	Significant
	Continuance Intention	H = 16.71	.002	Significant
Occupation	AI Usage Behavior	H = 36.83	<.001	Significant
	Perceived Value	H = 32.22	<.001	Significant
	Continuance Intention	H = 27.14	<.001	Significant
Overseas travel frequency	AI Usage Behavior	H = 139.77	<.001	Significant
	Perceived Value	H = 135.72	<.001	Significant
	Continuance Intention	H = 110.40	<.001	Significant
AI use frequency (pre-trip)	AI Usage Behavior	H = 191.58	<.001	Significant
	Perceived Value	H = 146.96	<.001	Significant
	Continuance Intention	H = 127.23	<.001	Significant

Legend: Significant at p-value < .05. U = Mann-Whitney U statistic; H = Kruskal-Wallis H statistic.

The strongest and most consistent differences were found for overseas travel frequency and pre-trip AI use frequency. Respondents who had traveled more than 10 times in the past three years reported the highest AI usage behavior (M = 4.59) and continuance intention (M = 4.67), whereas respondents with no recent overseas travel reported the lowest scores (M = 1.80 and 1.60, respectively). This pattern indicates that repeated travel experience may provide more opportunities for travelers to encounter situations in which AI can be useful, such as comparing destinations, organizing transportation, finding accommodations, and identifying restaurants or attractions. Frequent travelers may consequently have more opportunities to evaluate AI through actual use rather than through general perceptions of the technology.

A similar pattern was observed for pre-trip AI use frequency. Respondents who reported using AI “very often” before a trip recorded markedly higher perceived value (M = 4.75) and continuance intention (M = 4.76) than respondents who used AI less frequently. This finding suggests a close relationship between repeated interaction with AI and favorable evaluations of its usefulness. As travelers become more accustomed to incorporating AI into their planning routines, they may identify more situations in which the technology can save time, provide personalized recommendations, or simplify information searching.

Taken together, these findings indicate that AI usage behavior, perceived value, and continuance intention are closely associated with travelers’ actual travel and AI-use experience. Travelers who travel more frequently and use AI more frequently tend to report higher perceived value and stronger intentions to continue using AI. At the same time, these results should be interpreted as statistical associations rather than evidence of a causal relationship. Overall, the findings highlight the importance of actual usage experience in understanding AI adoption in travel planning and suggest that continued interaction with AI may be accompanied by increasingly favorable evaluations of its practical value.

3.5. Relationship Between AI Usage Behavior, Perceived Value, and Continuance Intention

All three relationships were positive and highly significant, indicating that the three major constructs examined in this study are closely interconnected. AI usage behavior and perceived value showed a very strong positive correlation (ρ = .826), indicating that respondents who reported using AI more extensively across different travel-planning tasks also tended to evaluate its benefits more favorably. This relationship may reflect the practical value that travelers obtain through repeated interaction with AI. As travelers use AI to search for destinations, organize itineraries, find accommodations and transportation, and obtain restaurant or attraction recommendations, they may become more aware of its usefulness, convenience, and personalization. Conversely, it is also plausible that travelers who already perceive AI as useful and trustworthy are more willing to incorporate it into a greater number of planning activities. Because the study employed a correlational design, the direction of this relationship cannot be established from the findings alone.

Perceived value showed the strongest relationship with continuance intention (ρ = .740), slightly exceeding the relationship between AI usage behavior and continuance intention (ρ = .679). This indicates that favorable evaluations of AI are particularly closely associated with travelers’ willingness to continue using it for future travel planning. In other words, actual use of AI appears to be important, but how travelers evaluate the benefits they receive from that use may be even more closely related to their intention to continue. The result is consistent with the general logic of the Technology Acceptance Model (Davis, 1989), which emphasizes perceived usefulness as an important factor in technology acceptance and use. In the context of travel planning, this perceived value encompasses not only usefulness but also personalization, convenience, time savings, and trust in the information provided.

The correlation between AI usage behavior and continuance intention (ρ = .679) was also strong and statistically significant, suggesting that travelers who use AI more frequently and across more planning activities tend to report stronger intentions to use it again. Repeated use may allow travelers to become familiar with AI’s capabilities and develop established habits around incorporating AI into their travel-planning process. However, the somewhat stronger association between perceived value and continuance intention suggests that frequency of use alone may not be sufficient to sustain future adoption. Travelers may continue using AI when their previous interactions have produced benefits that they consider meaningful and reliable.

Table 5. Relationship between AI usage behavior, perceived value, and continuance intention. 

Variable Pair	rho	p-value	Interpretation
AI Usage Behavior ↔ Perceived Value	.826	<.001	Highly Significant
AI Usage Behavior ↔ Continuance Intention	.679	<.001	Highly Significant
Perceived Value ↔ Continuance Intention	.740	<.001	Highly Significant

Legend: Significant at p-value < .05.

Table 5. Relationship between AI usage behavior, perceived value, and continuance intention. 

Variable Pair	rho	p-value	Interpretation
AI Usage Behavior ↔ Perceived Value	.826	<.001	Highly Significant
AI Usage Behavior ↔ Continuance Intention	.679	<.001	Highly Significant
Perceived Value ↔ Continuance Intention	.740	<.001	Highly Significant

Legend: Significant at p-value < .05.

Taken together, these findings suggest a closely connected pattern in which AI usage, perceived value, and continuance intention are positively associated with one another. The results therefore support the importance of moving beyond simple measures of whether travelers have used AI and examining how they evaluate the experience after using it. From a practical perspective, travel technology providers seeking to build long-term engagement should prioritize the quality, accuracy, relevance, and personalization of AI-generated outputs rather than simply expanding the number of tasks that AI can perform. Improving the usefulness and credibility of recommendations may contribute to stronger perceived value, which is closely associated with travelers’ willingness to continue using AI in future travel planning (Meng et al., 2026).

4. Conclusion and Recommendations

Most respondents were Korean travelers aged 45 and above, male, employed, and moderately frequent overseas travelers, and nearly all had used AI for a travel-related purpose. Respondents “agreed” that they use AI across all major trip-planning tasks, most notably for restaurant and attraction recommendations, suggesting that AI has already become a practical source of information for a range of travel decisions. Respondents also “agreed” that AI provides personalized, time-saving, convenient, and trustworthy travel information. At the same time, they expressed a comparable level of concern about the potential inaccuracy of AI-generated information. This finding suggests that travelers may recognize the practical advantages of AI while remaining aware that AI-generated recommendations should not always be accepted without verification, particularly when decisions involve important travel arrangements (Topsakal, 2025).

AI usage behavior, perceived value, and continuance intention did not differ significantly by gender but differed significantly by age, occupation, overseas travel frequency, and AI use frequency. In particular, older respondents, more frequent travelers, and more frequent AI users reported comparatively more favorable evaluations of AI. These differences suggest that travelers’ familiarity with travel planning and repeated exposure to AI may contribute to how they perceive the usefulness and value of AI-based travel services. The findings also indicate that AI adoption should not be viewed as a uniform process across all traveler groups. Differences in experience, travel frequency, and frequency of AI interaction may influence how travelers evaluate and incorporate AI into their planning behavior (Meng et al., 2025).

Perceived value emerged as the strongest correlate of continuance intention (ρ = 0.740, p < .001), followed closely by AI usage behavior (ρ = 0.679, p < .001). These statistically significant positive relationships indicate that respondents who evaluated AI more favorably and used AI more extensively were also more likely to report an intention to continue using it for future travel planning. The findings are consistent with the general logic of technology-acceptance theory, in which perceived usefulness and positive evaluations of a technology are associated with continued adoption. However, the results also demonstrate that favorable perceptions coexist with concerns regarding information accuracy, highlighting the importance of maintaining a balance between convenience and reliability in AI-supported tourism services (Lynna et al., 2025).

Based on these findings, several recommendations are proposed. First, travel technology providers and online travel agencies should continue investing in the personalization, usability, and factual accuracy of AI-generated recommendations. Because perceived value showed the strongest relationship with continuance intention, improving the quality of the information and ensuring that recommendations are relevant to individual travelers may strengthen users’ willingness to continue using AI. Systems should also clearly distinguish between verified information and AI-generated suggestions where appropriate, particularly for information that can change rapidly (Miguel et al., 2025).

Second, destination marketing organizations may consider AI-assisted planning tools as an additional communication channel for reaching experienced and higher-frequency travelers, who in this study showed comparatively strong AI usage and favorable evaluations. AI-based destination services could provide personalized recommendations for attractions, restaurants, transportation, and activities while helping travelers organize information according to their individual interests and schedules (Mladenović et al., 2026).

Third, targeted onboarding and simplified AI-planning features could be developed for younger and less experienced travelers who reported comparatively lower usage and perceived value despite the generally high level of digital accessibility in South Korea. Rather than assuming that digital familiarity automatically results in effective AI adoption, travel platforms could provide simple examples, guided prompts, and task-specific functions that demonstrate how AI can be used for practical travel-planning activities (Ruizhe et al., 2025).

Fourth, AI travel services should provide mechanisms that encourage users to verify information before making important decisions. Pricing, transportation schedules, opening hours, reservation availability, visa requirements, safety information, and other time-sensitive details can change frequently. Integrating links to authoritative sources, timestamps, or verification indicators may help travelers evaluate the reliability of AI-generated information while retaining the convenience of conversational planning (Molka, 2025).

Finally, future researchers should examine the specific sources of travelers’ concerns about AI accuracy through qualitative interviews, focus groups, or mixed-method research. Future studies could investigate which types of AI-generated information travelers consider most reliable and which types require human verification before booking or purchasing decisions are finalized. Longitudinal studies could also examine whether continued real-world use of AI changes travelers’ perceptions of usefulness, trust, and accuracy over time. Comparative studies involving travelers from different countries or cultural backgrounds may further determine whether the relationships identified in this study are specific to Korean travelers or reflect broader patterns of AI adoption in international tourism (Seyfi et al., 2026).

Overall, the findings indicate that AI has become a meaningful component of travel planning among the Korean travelers surveyed in this study. Continued adoption appears to be closely associated with the perceived value and actual use of AI, while concerns about information accuracy remain an important consideration. Strengthening personalization, convenience, transparency, and information credibility may therefore help travel-related AI services provide greater value while supporting more informed and responsible use of AI throughout the travel-planning process (Safari, 2026).