How AI chatbot service quality drives continuance intention: an S–O–R perspective in Saudi higher education

Abstract

Introduction:

Digital transformation is reshaping customer service, with AI-powered chatbots enhancing user interactions and becoming increasingly integrated into business operations. In higher education, rapid advancements in artificial intelligence have heightened interest in understanding the factors influencing students' continuance intention (CI) toward AI chatbot use. This study develops and validates an integrated model based on the Stimulus-Organism–Response (S-O-R) framework to explain students' CI toward AI chatbots in the Kingdom of Saudi Arabia.

Methods:

A quantitative research design was employed using convenience sampling to collect data from 370 students enrolled in three public universities in Saudi Arabia. The proposed model examined the influence of AI chatbot service quality (AICSQ) on students' continuance intention through perceived AI intelligence (PAI), perceived trust (PT), and user satisfaction. Structural Equation Modelling (SEM) using SPSS AMOS (Version 29) was applied to test the hypothesized relationships and sequential mediation effects.

Results:

The findings reveal that AICSQ, PAI, PT, and user satisfaction have significant direct effects on students' continuance intention to use AI chatbots. Furthermore, PAI, PT, and user satisfaction significantly mediate the relationship between AICSQ and continuance intention, both independently and sequentially. All proposed hypotheses were supported, while the control variables showed no significant influence on continuance intention.

Discussion:

The study highlights the critical role of AI chatbot service quality, perceived AI intelligence, perceived trust, and user satisfaction in fostering sustained chatbot usage among university students. These findings provide valuable practical insights for higher education institutions seeking to enhance AI chatbot adoption and long-term engagement. The study also contributes to the existing literature by extending the S-O-R framework within the context of AI-enabled educational services and identifies limitations and future research directions for further investigation.

1 Introduction

Digital transformation has become a key driver in modern customer service, with technologies such as chatbots playing a vital role in enhancing user experiences and optimizing business processes (Ngo et al., 2025). A chatbot is intelligent, conversational software that accepts natural language input, whether in text, voice, or a combination of both. It outputs a human response and can be programmed to operate in automatic mode during the work process (Radziwill and Benton, 2017). The notion of Chatbot has been around for quite some time in today's fast-evolving tech-driven society (Ask et al., 2016). Their popularity increased with Apple's launch of Siri, a voice assistant, in 2011, leading to widespread CI (Johnson et al., 2012).. Recent advancements in Artificial Intelligence (AI), particularly in Natural Language Processing (NLP) and Machine Learning (ML), have further boosted interest in chatbot technology (Rahman et al., 2017).

The advancement of AI is enabling enterprises to use chat for customer service. Indeed, an increasing number of managers are interested in the possibility of using chatbot to automate tasks in customer relationship services (Sands et al., 2021; Sheehan et al., 2020), including e- commerce (Araújo and Casais, 2019; Sundjaja et al., 2025; Yang et al., 2025), education (Mohammed et al., 2025; Polyportis, 2024), hospitality (Calvaresi et al., 2023; Pillai and Sivathanu, 2020; Seo and Lee, 2021), fashion (Murtarelli et al., 2023), health (Athota et al., 2020; Chung and Park, 2019), banking and finance (Asquer and Krachkovskaya, 2022; Feng et al., 2025; Graham et al., 2025; Huang, 2025). Always on and capable of processing massive amounts of requests, chatbots are built to help students with their problems in a timely and effective manner. This contributes to the generation of positive student experiences in the sense that, by providing a rapid and convenient access channel to knowledge and support, Chatbot can improve student satisfaction and foster loyalty (Kuhail et al., 2023; Selamat and Windasari, 2021).

In 2030, the global chatbot market is projected to be worth $3.99 billion, according to the recent Research and Markets report, rising at a 25.7% rate from 2022 to 2030. The education chatbot market is anticipated to observe an CI rate of 30.8% during the period from 2020 to 2027 on account of increasing interest in messaging platforms and a growing shift towards a more personalized learning experience (Al-Abdullatif, 2023). Vision 2030 in Saudi Arabia underscores explicitly the need for progress in the education sector through a roadmap of global best practices that guarantees a significantly improved educational system upon implementation. AI is one field that sees a connection to the future of education, and further development is essential for reaching the Kingdom's long-term objectives (Al-Amri and Al-Abdullatif, 2024).

There has been a growing interest in the application of Industry 4.0 technologies in higher education, which receives considerable support and can be considered a good starting point for their implementation within higher education institutions (HEIs) (Ayanwale, 2024). AI in education has developed into a dynamic academic discipline, enhancing the capabilities of not only learners but also educators and educational institutions (Stöhr et al., 2024). Students commonly utilize AI chatbots for academic purposes, such as tutoring, language learning, essay writing, and summarising research, enabling them to personalize their learning based on individual needs, preferences, and learning pace. With AI-driven writing assistance tools that offer on-the-spot grammar, spell check, punctuation, style suggestions, or even rewrites for better sentence structure, word choice, or tone, students can improve their writing at large. AI Chatbot can be provided as an individual self-study tool, allowing students to quickly obtain information and receive answers to their questions or issues in real-time (Malmström et al., 2023).

Students often see these tools as necessary for their future careers and are therefore supported in higher education. That's why exploring chatbots is a significant topic of discussion among higher education (HE) students. Understanding the human drivers that influence individuals’ decisions to adopt these technologies is crucial for developing effective methods to facilitate their uptake (Ayanwale, 2024). The introduction of tech-savvy tools in the education sector presents challenges that may require alternative solutions; yet, user Continuance Intention (CI, hereafter) remains a vital consideration in the Education system. This research paper therefore seeks to examine the factors that affect the CI of chatbots in an educational setting (Yildiz Durak and Onan, 2024). CI, hereafter, denotes a user's probability of utilizing, in the future, a given information technology system (Bhattacherjee, 2001).

While previous studies on the implementation of AI-based chatbots in education have primarily examined early user acceptance and CI, recent studies take a different approach, focusing on AI chatbot Service Quality (AICSQ, hereafter), perceived AI Intelligence (PAI, hereafter), perceived chatbot trust (PT, hereafter), chatbot satisfaction, and CI (Ashfaq et al., 2020; Avcılar and Yenilmez, 2026; Kavitha and Joshith, 2025; Lee et al., 2023; Li et al., 2021; Ngo et al., 2025). This new focus is important because AICSQ, PAI, PT, and satisfaction could contribute to increased student engagement and improved student performance, as well as improve institutional efficiency (Ahn, 2025; Ashfaq et al., 2020; Lee et al., 2023).

Research evidence suggests that service quality is one of the most important predictors of chatbot CI (Ashfaq et al., 2020). It is used to describe how students evaluate chatbot service offerings, such as the accuracy of responses, speed of replies, ease of use, level of personalization, and overall efficiency (Choi et al., 2025; Shahzad et al., 2024). Research shows that high-quality chatbot services improve chatbot CI (Ngo et al., 2025; Nguyen and Le, 2025; Shi et al., 2025), while poor service quality, such as low confirmation, unclear communication, inaccurate information, and slow responses, reduces satisfaction and trust, thereby weakening CI (Kavitha and Joshith, 2025; Pham and Ngo, 2026; Yu, 2023). Moreover, recent research revealed that service quality is a strong predictor of PAI (Chen et al., 2022). PAI is a term that describes how much users think that AI tools like Chatbot are intelligent, problem-solving, and cognitively able (Lee et al., 2023). Again, PAI is a strong indicator of PT (Lu et al., 2025). PT in AI refers to the impressions held by users about how reliable, reputable, and dependable AI technologies are (Baroni et al., 2022). PT is key for the CI of technology, as users have to feel confident and secure in the system's ability and ethical use (Choudhury and Shamszare, 2023; Dahri et al., 2024). Additionally, PT is a forecaster of satisfaction (Silva et al., 2023; Wut et al., 2025). Satisfaction refers to the positive emotional response users experience after evaluating how well the chatbot has performed its tasks (Bhattacherjee, 2001). It is one of the most important predictors of chatbot CI (Avcılar and Yenilmez, 2026; Kavitha and Joshith, 2025; Ngo et al., 2025).

However, how AI Chatbot Service Quality (AICSQ) forms CI through Perceived AI Intelligence (PAI, hereafter), Perceived Trust (PT, hereafter) in Chatbot, satisfaction remains inadequately understood (Chotisarn and Phuthong, 2025; Li et al., 2021; Nguyen et al., 2021).

Figure 1

This article embodies several values relevant to the research on Chatbot. First, it highlights a new and relatively understudied area: the factors that influence students’ CI to use Chatbot. This adds more depth to the current literature about chatbot acceptance systems. Secondly, the study employs an in-depth model to illustrate what motivates users to continue using a chatbot. It extends the S-O-R model (Figure 1) by integrating AICSQ, PAI, PT, and satisfaction, providing a more comprehensive explanation of students’ chatbot reuse decisions. This model has been tested for its strength and utility and was found to provide practical appraisal as a reference for future research inquiries. Third, the study draws on data from students with prior experience with chatbots and highlights strategies to promote sustainable chatbot use. Its findings can guide quality assurance specialists, e-service providers, software developers, academic heads, and e-learning technology managers in improving chatbot CI and ensuring sustainable usage within the education sector.

2 Theoretical frameworks and existing literature

2.1 The stimulus-organism-response (S-O-R) framework

The S-O-R paradigm was proposed by environmental psychologists Mehrabian and Russell (Mehrabian and Russell, 1974) and has since been widely used to study the human decision- making process (Kim and Lennon, 2013; Rafiq et al., 2022). The framework of S-O-R implies that factors external to or cues coming from the environment are stimuli (S) that form the internal psychological state of a person (O), which then lead them to corresponding behavioral actions or reactions (R) (Alzahrani and Bhunia, 2024; Anwar et al., 2023; Chopdar and Balakrishnan, 2020). Under this framework, stimuli can be the factors of the environment or the level of information presented. The organism is the internal state of an individual, including their emotional reactions to these stimuli. The response, as the final component, follows from those behaviors that could be directed at something or somebody, or from withdrawing oneself from a situation (Alzahrani and Bhunia, 2024; Çera et al., 2022).

In this research, S is identified as AICSQ, as it refers to the external qualities of the service that students encounter when communicating with AI Chatbot. O consists of PAI, PT in Chatbot, and Satisfaction, as it concerns students’ internal psychological processes regarding their experience with the chatbot service. On the other hand, the R is the CI, as it reflects students’ behavioral reaction to using AI Chatbot continuously in education.

This research also demonstrates how the stimulus and organism variables directly influence the response; their indirect and combined effects through mediations and serial mediation are also considered.

2.2 AICSQ and CI to use chatbot

Interactions between chatbots tend to focus more on the task at hand and are generally shorter than those between users and human beings (Turpening and Littleton, 2016). They do not feel tired like human beings (Ruan and Mezei, 2022) and are gaining increasing popularity by providing fast self-service solutions supported by continuous improvement in AI, NLP, ML, and chatbot technologies (Rahman et al., 2017). In contrast to basic automated responders that generate text or speech, the dynamic of human–machine communication is what chatbots are designed for (Prasetya et al., 2018). In such dialogues, these interactions commonly take the form of a user question or command within brief, goal-oriented exchanges (McTear et al., 2016). Reasonably current Chatbot also utilize context awareness (Pearl, 2016), which enables them to detect when the user has stopped talking, interpret messages correctly, infer meaning beyond the literal text, recognize linguistic cues (e.g., politeness), prevent further problems or misunderstandings, and more (Li et al., 2024; Skantze, 2007).

In terms of AI Chatbot assessment, students can notice some principal characteristics. The first one is understandability, which shows how correctly a chatbot based on NLP techniques interprets user inputs (Lu et al., 2024; Mehrolia et al., 2023). The second is reliability supported by algorithmic systems, which guarantee consistent and accurate responses (Hassani and Silva, 2023; Li et al., 2021). The third is responsiveness, which refers to how quickly a chatbot can provide students with the answers they need to obtain services, thereby enhancing the overall experience (Mehrolia et al., 2023). Quality number four is assurance, which concerns users’ trust in the accuracy and safety of the provided information, reinforced by clear communication, privacy compliance, and the delivery of exact answers (Li et al., 2021; Rajaobelina et al., 2021). Lastly, interaction refers to the chatbot's ability to engage in natural conversation, enabling more effective user–system relationships for various purposes (Li et al., 2021; Li and Zhang, 2023; Sundjaja et al., 2025). Research indicates that higher-quality chatbot services are associated with greater chatbot CI (Han et al., 2025; Manigandan and Alur, 2024; Paraskevi et al., 2023; Shahzad et al., 2024). On the other hand, service quality inadequacy, such as lower confirmation, poor understandability, incorrect information, or slow responses, repeated inaccuracies, negatively affects chatbot CI, which is crucial for user satisfaction and thus adversely affects it. Failure in comprehensibility, accuracy, and interaction delay results in unmet expectations and, consequently, lower satisfaction, thereby weakening chatbot CI. In addition, a mismatch between information quality and service quality, that is, when there is asymmetry between the two levels of performance, negatively impacts users’ trust and chatbot CI more than balanced performance levels do (Kavitha and Joshith, 2025; Pham and Ngo, 2026; Yu, 2023). Based on the above discussion, the following hypothesis is proposed:

H1: AICSQ will have a significant positive effect on students’ chatbot CI. 2.3 PAI and CI to use chatbot

The intelligence displayed by machines, rather than by humans, is referred to as Artificial Intelligence (AI). Intelligence, as demonstrated by humans or animals, involves consciousness and emotions, whereas the latter lacks these attributes. The term AI was first used by John MsCarthy in 1955, and he defined it as “making a machine behave in ways that would be called intelligent if a human were so behaving” (Shannon, 1955). Alan Turing popularized the idea that a computing machine may think like humans someday in his research work (Turing, 1950.). In the future, automated machines would calculate things that no one would rationally do today, he believed.

The 21st century has witnessed the digital transformation of HEIs, a trend that accelerated during the previous decade (Al-Tahitah et al., 2021).. In the era of AI, higher education institutions are increasingly utilizing chatbots for instructional purposes, offering quick and personalized support to all stakeholders, including students and staff (Al-Sharafi et al., 2023). AI intelligence is typically characterized by its capability for problem-solving, learning, and self-adaptation over time, as well as the ability to manage information and provide appropriate responses to various environments (Legg and Hutter, 2007).

In the education sector, AI chatbots are increasingly recognized as interactive teaching tools that can deliver content, foster independent learning, engage learners, and provide immediate feedback, supporting personalized learning and enhancing long-term memory (Jeon, 2024; Mageira et al., 2022; Zhang et al. 2024a). Complementing human instructors, they provide ongoing support, particularly valuable in online learning and language education, which can reduce language anxiety, boost student engagement, support self-directed learning, improve information retention, and promote continuous, multimodal learning experiences (Mohebbi, 2025; Shawar, 2017).

The interaction between human users and educational chatbots plays a crucial role in promoting learner engagement and improving academic productivity (Akpan et al., 2025). The more a robot or chatbot resembles a human being or a living thing, the more intelligent it appears to its users (Bartneck et al., 2009). Therefore, when users perceive chatbot as genuinely helpful and competent, they are more likely to apply and incorporate it into their learning activities (Tian and Wang, 2022). In this regard, the PAI has significant implications for metacognitive self-regulated learning. Consequently, the PAI results in a positive influence on the intention to use Chatbot (Balaskas et al., 2025; Dahri et al., 2024; Shahzad et al., 2024; Shannon, 1955). Based on the earlier points, we suggest the following hypothesis:

H2: PAI will have a significant positive effect on students’ chatbot CI. 2.4 PT and CI to use chatbot

Although trust has been studied across marketing, psychology, and information systems, the puzzle posed by its definition prevents a universal definition (Rahim et al., 2022). It is what depicts an individual's confidence in the truthfulness and reliability of a system that promotes technology use, thereby reducing perceived risk (Guo and Erdenebold, 2025).. This is especially true for AI technologies, primarily chatbots, as they are cutting-edge and still relatively unknown (Lim et al., 2023). In the present paper, PT refers to students’ faith in the quality, reliability, security, and overall performance of AI services in higher education, encompassing openness and trustworthiness regarding privacy and chatbot responsiveness, as well as other software features that foster trust (Guo and Erdenebold, 2025).

Early trust is crucial, as many students are still coming to grips with AI-based conversation agents, and it cannot be said that these systems are transparent about their intentions (Kaabachi et al., 2019; Kim et al., 2011). Trust promotes the use of chatbots as it gives the impression of reliability (Hassani and Silva, 2023; Lu et al., 2024; Mehrolia et al., 2023) and helps people to form closer relationships with digital devices (Jyothsna et al., 2024). It is acquired when individuals feel they will not be taken advantage of, even in uncertain online environments (Huang and Lee, 2022; Mehrolia et al., 2023), and serves as an essential factor in both human-based and technology-mediated interactions (Nguyen et al., 2021). People may be hesitant to share personal information when interacting with a chatbot (Choudhury and Shamszare, 2023) as trust is often low, which consequently decreases acceptance and satisfaction with chatbot technologies, ultimately leading to client rejection (Mostafa and Kasamani, 2022).

There is no lack of evidence regarding the reliance on operating such robots within customer care, particularly within the health sector, and further within the banking industry, as shown in past works (Devisakti and Muftahu, 2023; Kasilingam, 2020; Sitthipon et al., 2022), with a greater trustworthiness and a majority of customers’ willingness to demonstrate the use of these tools (Gatzioufa and Saprikis, 2022). However, although AI chatbots are widely used in consumer services, empirical evidence in the higher education setting remains limited. In education, interest in studying chatbot acceptance has grown in recent years (Kooli, 2023). However, most of this work has been qualitative, and there remains a significant shortage of empirical studies (Almufarreh, 2024). Studies on chatbot CI among university students found that PT has a direct positive effect on students’ intentions to use Chatbot for learning (Guo and Erdenebold, 2025; Kuberkar and Singhal, 2020; Rahim et al., 2022). From the above discussion, we can put forward the following hypothesis:

H3: PT will have a significant positive effect on students’ Chatbot CI 2.5 Satisfaction and CI to use chatbot

User satisfaction can be described as the way consumers evaluate a product or service after they have used it (Kim, 2019). Satisfaction is often used to indicate how well a product or service meets customer expectations in business (Chung et al., 2020), the overall sense of pleasure that builds up over time through repeated exposure to media (LaRose, 2010). It helps show whether a system is successful and is affected by users’ experiences and the system's results (Urbach and Müller, 2011). This satisfaction grows over time, as more interactions are created. In technology acceptance research, satisfaction plays a significant role, as it directly influences users’ future intentions and their likelihood of continued use of the system (Ashfaq et al., 2020; Yu et al., 2024).

By using digital platforms, chatbots can provide information clearly and precisely, significantly reducing users’ confusion and contributing to their overall satisfaction (Chung et al., 2020.; Mimoun et al., 2017). They can achieve this by tailoring their communication to each user's profile, thereby further boosting the overall service experience (Ferreira and Barbosa, 2023; Silva and Canedo, 2024). Research has consistently led to the development of chatbots that are trustworthy, personalized, up-to-date, and reliable, resulting in consistently positive customer satisfaction. Studies have shown that user satisfaction plays a key role in shaping Students’ CI to use chatbot (Ashfaq et al., 2020; Cheng and Jiang, 2020; Hsiao and Chen, 2022; Hsu and Lin, 2023; Huang and Chueh, 2021; Jang et al., 2023; Nguyen et al., 2021). Based on the discussion, the following hypothesis can be proposed.

H4: Satisfaction will have a significant positive effect on students’ chatbot CI

2.6 Mediating role of PAI

AI chatbot services significantly modify users’ attitudes towards AI, with service quality directly influencing perceptions of fulfillment, assurance, and competence. On a one-to-one basis, this approach makes conversations more personalized and culminates in an enhanced perception of intelligence (Acayıp, 2024), Rapid, accurate responses better serve consumers and reinforce the chatbot's functions (Chen et al., 2022). Intelligent communication, as well as non-verbal characteristics, will help define intelligence in users’ opinions (Wuenderlich and Paluch, 2017), thereby supporting the emotional and cognitive credibility required for PAI enhancement. On the other hand, when chatbots provide poor service or give incorrect information, people may perceive them as less intelligent or as not deserving their trust, thereby prompting calls for improvement (Chen et al., 2023). Moreover, AI-powered Chatbots could exhibit intelligence by delivering appropriate dialogue in context and retaining conversational context. The consistency with which relevant, immediate, personalized responses affect current applications of AI technologies, shaping consumer attitudes, is widely recognized by researchers (Bernabei et al., 2023; Sohail et al., 2023). Perceived intelligence plays a significant role in the acceptance/usage of AI innovations. This factor is delivered through students’ awareness of the chatbot's capabilities or limitations (Abdelkader, 2023; Fui-Hoon Nah et al., 2023; Sahari et al., 2023). When users perceive the chatbot as intelligent, they are more likely to use it (Balaskas et al., 2025; Dahri et al., 2024; Shahzad et al., 2024; Shannon, 1955). Based on this discussion, this study suggests that:

H5: PAI will mediate between AICSQ and the students’ chatbot CI. 2.7 Mediating role of PT

Service quality is the extent to which information systems meet users’ needs through accurate, timely, and knowledgeable responses; it is also the gap between what users expect from a service and their actual service experience (Trawnih et al., 2022). Overall system quality, including information, system performance, and service delivery, influences user satisfaction and continued use (Chaw and Tang, 2018; Mehrolia et al., 2021; Mouakket, 2020). Users expect to access trustworthy online platforms that provide helpful, up-to-date, and relevant information (Alagarsamy and Mehrolia, 2023; Kaabachi et al., 2019; Khan et al., 2017; Leonard and Jones, 2021; McKnight et al., 2002; Mehrolia et al., 2021; Roca et al., 2006; Trivedi, 2019; Yoon and Occena, 2015). Trust in e-commerce and m-commerce app-based services is anchored in information quality as a core driver, even more than in search engine speed or accuracy. Consistency in reliability, responsiveness, assurance, and personalization will further develop trust in the service offered (Chung et al., 2020; Li et al., 2021). Although interaction promotes engagement, trust mostly counts on users’ quality judgments (Nilashi et al., 2015). Smart device interactivity supports interface design, aesthetics, post-use trust development, and positive user attitude incrementing (Yang and Lee, 2017). Studies have found that AICSQ has a strong positive effect on trust (Alagarsamy and Mehrolia, 2023; Nguyen et al., 2021; Trawnih et al., 2022).

Moreover, trust is particularly significant when it comes to AI tools like LLM-based chatbots, for one reason: they are new and complex (Lim et al., 2023). PT was primarily used in this study to compare students’ confidence levels in the reliability, security, smartness, and overall performance of AI chatbot services in higher education. By interacting more like a human during conversations, users may also risk being impersonated by impostors, such as hackers, who create fake chatbot personas (Nguyen et al., 2021), thereby affecting their trust in or against using the application. Trust is based on the fact that the chatbot performs its function dependably; on how open it is with people; on how well it guards their personal data; and, finally, on how effectively it responds to people. Research on chatbot use in universities shows that when students trust chatbot, they are more likely to use them for learning (Guo and Erdenebold, 2025; Kuberkar and Singhal, 2020; Rahim et al., 2022). Based on this, we propose the following hypothesis:

H6: PT will act as a mediator between AICSQ and students’ chatbot CI. 2.8 Mediating Role of User Satisfaction

System quality is the primary factor in determining whether users encounter difficulties when interacting with AI chatbot, and it is essential to avoid technical problems and bugs. It is a high contributor to user satisfaction as well (Ngo et al., 2025). Having excellent system quality leads to better accuracy of the AI-generated recommendations, thus increasing user satisfaction (Kim and Kim, 2025). For example, if error rates are low and responses are quick, users experience little frustration, and their interactions are smoother than ever before. Research indicates that such factors as system compliance can significantly improve young customers’ satisfaction (Elayat and Elalfy, 2025). It was found that high system quality creates a positive impression on customers regarding chatbot services; therefore, its importance to the overall user experience is evident (Mulyono and Sfenrianto, 2022). These findings show that, for AI chatbot, reducing the effort users feel they must put into technical tasks is essential to creating smooth, effective interactions. Therefore, the quality of systems can be regarded as a basic factor affecting satisfaction in the usage scenario for AI chatbot (Ngo et al., 2025).

Among other measures, user satisfaction is considered a significant indicator of a system's success and reflects customers’ attitudes toward a product and its outcomes (Urbach and Müller, 2011). It grows gradually with time when the individual continuously interacts with that system. In technology CI and user behavior studies, satisfaction is indicated as a key determinant that greatly influences users’ future intentions towards a system (Ashfaq et al., 2020). A significant number of studies confirm that delighted customers are loyal in the long term and are among the key reasons for continued system use (Nascimento et al., 2018). Hence, if chatbot users feel the service meets their initial expectations, they become satisfied. When they remain confident over time, they are more likely to keep using the chatbot (Li et al., 2021; Silva and Canedo, 2024; Yu et al., 2024). Therefore, we put forward the following hypothesis:

H7: User Satisfaction will mediate between AICSQ and students’ Chatbot CI.

2.9 Sequential mediating effects of PAI, PT, and user satisfaction

High-quality AI chatbot services influence students’ perceptions by enhancing accuracy, responsiveness, and natural interactions, which boosts PAI (Acayıp, 2024; Wuenderlich and Paluch, 2017). When students see chatbot as intelligent, they are more inclined to adopt them (Balaskas et al., 2025; Dahri et al., 2024; Shahzad et al., 2024; Shannon, 1955). Higher PAI also increases PT (Lu et al., 2025), which is crucial because AI-based chatbot remain new and complex for many students. Trust alleviates uncertainty and boosts users’ confidence in the reliability, privacy, and performance of chatbot services (Luo et al., 2024; Nguyen et al., 2021; Saldivar Espejo et al., 2025; Wut et al., 2025). Trust subsequently enhances user satisfaction, as students perceive the system as meeting expectations and offering clear, reliable support (Nagashree et al., 2025). Previous research confirms that satisfaction strongly influences continued usage intentions (Ashfaq et al., 2020; Li et al., 2021; Luo et al., 2024; Nguyen et al., 2021; Silva and Canedo, 2024; Yu et al., 2024). Therefore, AICSQ may indirectly shape users’ intention to use chatbot through a chain reaction: higher PAI results in greater PT, which boosts satisfaction, ultimately reinforcing CI. Based on this reasoning, the following hypothesis is proposed:

H8: The effect of AICSQ on students’ chatbot CI is sequentially mediated by PAI, PT, and Satisfaction.

3 Methods

3.1 Research design

The survey was conducted as a cross-sectional study using a quantitative approach to empirically test the proposed theoretical model. Following the conceptual model, this research has employed the S-O-R framework, in which AICSQ and PAI are the stimulation components; PT and satisfaction are the organism variables; and chatbot CI is the behavioral response. In addition, it provides direction on how stimulus and organism variables directly and indirectly affect the response, and on their combined effects through mediations and serial mediation.

3.2 Sample and data collection

The sample group for this research comprises students from three universities in the Kingdom of Saudi Arabia: King Saud University, King Abdulaziz University, and the University of Jeddah. Data were collected through convenience sampling, a non-probability sampling method that is easy and fast to implement and cost-effective, making it ideal for exploratory research (Acharya et al., 2013; Ngo et al., 2025). The online questionnaire was developed in Google Forms and sent to participants via email in May 2025; reminders were sent until October 2025. In the email, the study's objective was detailed, chatbots were briefly introduced (education and general-purpose AI chatbots), and respondents’ perspectives were given significance. The participants’ identities remained anonymous; no personal questions were asked, and their responses were used solely for academic purposes. To reduce non-response bias, the survey was made mobile-friendly and kept short (7–8 min) (Bilquise et al., 2024).

The survey was completed by 405 individuals, and the data screening process was conducted to clean and prepare the data for further statistical analysis. After eliminating questionnaires with missing values and unengaged responses, a final usable sample of 370 respondents was retained for the study. Following Benitez et al. (2020), the sample size was determined using Cohen's statistical power analysis to ensure reliable and conclusive results (Baroudi and Orlikowski, 1989). This approach is strongly recommended for structural model analysis (Kock and Hadaya, 2018). The G*Power tool was used to conduct the power analysis, with a medium effect size (f² = 0.15) selected based on Cohen (2013) guidelines. Using α= 0.05, statistical power = 0.80, and 4 predictors, the analysis indicated that a minimum sample size of 85 participants was required (Juita et al., 2026). Moreover, earlier studies have recommended a minimum of 200 responses for the structural equation modelling (SEM) (Alzahrani and Bhunia, 2024; Alzahrani and Bhunia, 2025a; Bhunia and Shome, 2023; Boomsma, 1983; Hair et al., 2019, b) Further, Nicolaou and Masoner (2013) (Nicolaou and Masoner, 2013) suggest using 5 to 10 responses per measurement instrument for SEM analysis. Since the questionnaire consisted of 19 items, the final sample of 370 was far more than adequate and met all the criteria that were recommended in the literature. The mean age of the participants was 21.035 years. Table 1 shows the demographic profile of the participants in this study.

The information presented in Table 1 demonstrates a varied demographic composition of the survey respondents. In terms of gender distribution, the sample is predominantly male, with 240 respondents (64.900%), while females account for 130 respondents (35.100%). Regarding age, most participants fall within the 19–21 years age group, representing 255 respondents (68.900%). This is followed by 113 respondents (30.500%) aged 22–24 years, whereas only two respondents are 25 years old. Regarding employment status, the majority of participants, 323 respondents (87.300%), are full-time students. Additionally, 26 respondents (7.000%) reported holding part-time jobs, 10 respondents (2.700%) were engaged in on-campus employment, and 11 respondents (3.000%) were employed full-time outside their academic activities. Regarding enrollment status, seniors constitute the largest group, with 155 respondents (41.900%), followed by sophomores with 87 respondents (26.200%), juniors with 97 respondents (23.500%), and freshmen with 31 respondents (8.400%). Regarding student classification, 30 respondents (8.100%) are international students, while the majority, 340 respondents (91.900%), are domestic students. Furthermore, analysis of monthly family income reveals that 11 participants (3.000%) reported earnings below 2,000 SR. Smaller proportions fall within the 2,000–5,999 SR range (13 respondents, 3.500%) and the 6,000–9,999 SR category (37 respondents, 10.000%). Additionally, 62 respondents (16.800%) reported incomes between 10,000 and 15,000 SR, while the largest segment, 247 respondents (66.800%), reported monthly family incomes exceeding 15,000 SR.

3.3 Research instrument

An instrument with 19 items was developed to assess the constructs in the research model following a careful review of the literature. Two parts were present in the instrument demographic information and construct evaluation. The first part was primarily designed to collect demographic information, such as gender and age. The second section contained 19 items spanning five constructs in the proposed model, all adapted from prior, validated studies.

In-depth, AICSQ was assessed using four items adapted from Zhang et al. (2022). PAI was measured using four items adapted from Moussawi and Koufaris (2019). Additionally, PT was evaluated using four items based on the work of Gefen et al. (2003). User satisfaction was measured using four items adapted from Teo et al. (2008). Finally, CI was assessed using three items adapted from Bhattacherjee (2001). Responses to all items were recorded using a five-point Likert scale, with scores from 1 indicating “strongly disagree” to 5 indicating “strongly agree”. Appendix A outlines the constructs, their corresponding sources, and related items.

3.4 Control variables

The study under consideration includes control variables to eliminate potential confounders and ensure the reliability of causal inferences (Nielsen and Raswant, 2018). Since the independent variables might include demographic factors, many times-examined factors such as age, gender, employment status, enrollment status, student type, and family income are controlled for to ensure that the observed outcomes are not due to demographic variability. These are included in this research model, in line with previous academic literature, to improve accuracy when testing for independent variable effects (Ashfaq et al., 2020; Cheng and Mitomo, 2017; Fang et al., 2014; Li et al., 2020; Zafar et al., 2021).

3.5 Data screening

The collected data were subjected to a meticulous review to determine whether they were fit for further analysis. The data were then entered into SPSS 29.0 and AMOS 29 for analysis. At the first step, the data were checked for missing values and for surveys that were not completed correctly. It was found that the highest amount of missing information was at 7%, which is below the acceptable level of 10% for a variable on a specific item (Alzahrani and Bhunia, 2024; Bhunia and Shome, 2023; Alzahrani and Bhunia, 2025a, b; Cohen et al., 2013; Kline, 2023; Roy et al., 2017).

However, the missing data were handled using ‘Regression Imputation’. This procedure was chosen explicitly because Likert-type scale data were used in the current study (Lynch, 2003) and were supported by SPSS software version 29 (Final sample size 370). Normality testing is another prerequisite before SEM (Byrne, 2013). The dataset reflected a normally distributed sample, as indicated by skewness and kurtosis values within ±2, an acceptable range for each measurement item (Hair et al., 2019).

Given that a single instrument was used to measure all constructs, it is plausible that common- method bias may have influenced the dataset (MacKenzie and Podsakoff, 2012). The issue was examined using a single-factor test (Harman, 1976). When all items were forced into a single factor, it was found that this factor could explain 31.851% of the variance, which is less than 50%. Therefore, there is no reason to believe the dataset is biased.

The data were later analysed using AMOS (Analysis of Moment Structures) version 29, which was grounded in the covariance-based SEM (Roy et al., 2017). The reason for applying SEM is its ability to estimate multiple path relationships simultaneously. Furthermore, among other analysis methods, SEM has certain benefits, as it combines multiple regression with factorial analysis and path analysis, and treats measurement error directly (Hair et al., 2019; Trivedi and Pattusamy, 2022). The study took a two-step approach as proposed by Anderson and Gerbing (1988). Initially, based on the recommendation above, a measurement model was developed to assess whether the construct under study is actually represented by its indicators. Secondly, a structural model tests whether the hypothesized relationships are supported by empirical evidence.

4 Results

4.1 Measurement model

The measurement model assesses the validity and reliability of a construct included in a study (Henseler et al., 2009). Firstly, we conducted a Confirmatory Factor Analysis (CFA) to examine the hypothesized correlation structure among the variables (Figure 2). Their factor loadings checked the reliability of individual items. It was found that five items had factor loadings below the required threshold of 0.6: one item each from AICSQ (AICSQ_1), PAI (PAI_4), and PT (PT_1), and two items from SAT (SAT_1 and SAT_3). So, we deleted these items (Chin et al., 1997). After the removal of items from the SAT construct, it was measured using two items. However, this does not pose a methodological concern, as constructs measured with two items are considered acceptable when the items demonstrate strong convergent validity (Usakli and Rasoolimanesh, 2023). The results of our study confirm adequate convergent validity for the retained SAT items. The final CFA model showed a better fit (Table 2).

Figure 2

For the evaluation of construct reliability, both Composite Reliability (CR) and Cronbach's alpha (α) were employed. A value of 0.6 or higher is considered acceptable for composite reliability (Bagozzi and Yi, 1988), whereas for Cronbach's alpha, values of 0.7 or higher are considered adequate (Gefen et al., 2000). In this study, values of CR ranged from 0.799 to 0.863, while values of Cronbach's alpha ranged from 0.797 to 0.861 (Table 3). Therefore, it can be inferred that all constructs and their dimensions demonstrated reliability.

Furthermore, the research surveyed convergent and discriminant validity. For convergent validity, the Average Variance Extracted (AVE) was checked. All constructs had AVEs greater than 0.5 (Table 3), confirming convergent validity (Elnadi and Gheith, 2023). Furthermore, the Fornell–Larcker criterion was also used to assess Discriminant Validity (Fornell and Larcker, 1981). This criterion requires that the square root of the AVE for each construct exceeds its highest correlation with other constructs, as shown in the diagonal entries of Table 4. Moreover, all HTMT values fall below the 0.85 threshold, further confirming that discriminant validity has been adequately established (Henseler et al., 2015) (Table 5).

4.2 Testing hypotheses through the structural model

To assess the importance of the hypothesized pathways and support our model's predictions, we conducted a structural model. All fit indices for the proposed structural model indicate a perfect fit (Table 6). Hypotheses were examined by including predicted paths in the measurement model. The maximum likelihood estimation technique (MLE) was used to analyse the hypothesized paths. MLE is a commonly used estimation procedure that is known to produce accurate results under good conditions (Hair et al., 2019; Trivedi and Pattusamy, 2022).

4.2.1 Direct effects

Based on the proposed objectives and hypotheses, we constructed a path model with a beta weight (β) used to indicate the strength of the relationship, and then we evaluated it. The statistical significance was considered to be maintained when the p -value was at or below 0.05 (Biswas and Verma, 2021).

To fulfil the first objective, the study examined the direct effects of AICSQ, PAI, PT, and satisfaction on users’ CI to use chatbot (H1, H2, H3, and H4).

In the results, it was shown that CI to use chatbot was significantly influenced by AICSQ (β = 0.147, p = 0.020), PAI also had a significant effect on CI to use Chatbot (β = 0.151, p = 0.014), PT influenced substantially CI to use chatbot (β = 0.208, p = 0.001) and satisfaction was found to have a significant effect on CI to use chatbot (β = 0.288, p < 0.001). Therefore, hypotheses H1, H2, H3, and H4 were supported in this study. Furthermore, it was found that control variables did not have a significant effect on users’ CI to use chatbot.

4.2.2 Indirect effects

To fulfill the other objectives, mediation and serial mediation analysis were performed. Mediation analysis, a widely used statistical technique, examines how one variable affects another through intervening variables. Furthermore, SEM is considered a practical approach to verify this type of mediating process (Little et al., 2007). In this study, we employ a bootstrapping-based mediation mechanism to assess the significance of the indirect pathways that influence the outcome variable. 5000 bootstrap resamples were also used in the research, along with a 95% bias-corrected confidence interval, to determine whether the indirect effect is significant. This approach was selected due to its ability to produce more accurate confidence intervals and its higher statistical power (MacKinnon et al., 2004)

Regarding the second objective (H5), we found a distinct indirect effect of AICSQ on users’ CI to use chatbot, via PAI (β = 0.107, p < 0.001). Subsequently, in the quest for the third objective (H6), it was found that PT positively mediates the relationship between AICSQ and users’ CI in using chatbot (β = 0.050, p = 0.002). Also, for the fourth objective (H7), the indirect effect of AICSQ on users’ CI in using chatbot through user satisfaction was significant (β = 0.048, p = 0.007). Finally, regarding the fifth objective (H8), when examining AICSQ and users’ CI regarding chatbot use through the sequential mediation of PAI, PT, and satisfaction, we found that this was significant (β = 0.010, p < 0.001).

In all instances, the direct effect of AICSQ on users’ CI to use Chatbot, in the presence of mediators, was significant (β = 0.147, p = 0.020), indicating partial and serial mediation. Therefore, H5, H6, H7 and H8 were supported. The summary of the structural model results and the path diagram are provided in Table 7 and Figure 3.

Figure 3

5 Discussion

Due to the rapid development of AI in higher education, there is growing interest among the educational community in what motivates students to adopt emerging AI Chatbots. The role of these tools is exciting, as they provide scalable, personalized learning support; however, their impact depends on whether students are ready to engage with them (Balaskas et al., 2025). The present research developed an integrated model grounded in the S-O-R framework to test the impact of AICSQ on students’ chatbot CI in the Saudi Arabian context. Moreover, it examined the sequential mediation process among AICSQ, PAI, PT, user satisfaction, and CI in the use of chatbot. The results indicated that all direct and indirect paths in the proposed model are statistically significant, which means that all hypotheses are supported. The validated model suggests that AICSQ, PAI, PT, and user satisfaction are essential precursors of continuance intention towards chatbot use. In addition, the proposed model demonstrates moderate explanatory power by accounting for 34.400% of the variance in continuance intention (CI) toward using chatbots, as indicated by the adjusted R² value. This suggests that the independent variables included in the model collectively explain a substantial proportion of users’ intention to continue using chatbot technology. Therefore, the adjusted R² result confirms that the model possesses acceptable predictive capability and provides meaningful insight into the determinants of chatbot CI.

The significant effect of AICSQ on continuance intention to use chatbot is in line with earlier studies (Han et al., 2025; Manigandan and Alur, 2024; Paraskevi et al., 2023; Shahzad et al., 2024). chatbot service quality enables positive attitudes by fostering ease of use, reliability, and high- level performance that meet users’ expectations. Even though previous studies revealed system quality as one of the main determinants of users’ CI service robots, Magno and Dossena (2023) failed to detect an impact of system quality on consumers’ attitudes towards chatbots used by brands. The differences can be attributed to cultural differences between the countries under study, their levels of economic development, and demographic variables. Users of chatbots in transitional economies who belong to Generation Z are technologically savvy and pay particular attention to efficiency and responsiveness when using technology. Therefore, in technologically developing countries and amongst young consumers, performance issues become much more important for chatbot CI (Ngo et al., 2025).

Furthermore, results also emphasize a significant link between PAI and university students’ chatbot CI, further supporting previous research findings (Balaskas et al., 2025; Dahri et al., 2024; Shahzad et al., 2024; Shannon, 1955), while this contradicts the findings from the earlier study (Maheshwari, 2024) that found an insignificant impact of PAI on students’ chatbot CI. According to the study's findings, people who find chatbots accurate, intelligent, and insightful are likely to be interested in using the technology, since it would mean the tool is highly relevant to fulfilling academic obligations. In this sense, the results confirm how important the perception of intelligence is in engaging people in using AI technology, despite its obvious limitations (Balaskas et al., 2025; Strzelecki, 2024; Tiwari et al., 2024).

Additionally, this study found a significant positive association between PT and chatbot CI, consistent with earlier research (Guo and Erdenebold, 2025; Kuberkar and Singhal, 2020; Rahim et al., 2022) and, contrary to earlier studies (Pal et al., 2022; Silva et al., 2023). Previous studies also support the idea that trust plays a key role in promoting the acceptance of technology, particularly in AI and automated systems (Ding and Najaf, 2024). Trust helps reduce concerns about data privacy, security, and the ethical use of AI. Participants considered trust an important factor in their decision-making, particularly due to concerns about plagiarism and the misuse of AI-generated content in academic work (Amin et al., 2025; Balaskas et al., 2025; Foroughi et al., 2025). At the same time, privacy and security concerns can undermine trust in chatbot, as they rely on user data and may pose security risks by storing and learning from it (Alagarsamy and Mehrolia, 2023).

Moreover, the study revealed a significant positive relationship between user satisfaction and chatbot CI, which is in line with earlier studies (Ashfaq et al., 2020; Nguyen et al., 2021; Hsiao and Chen, 2022; Huang and Chueh, 2021; Jang et al., 2023; Cheng and Jiang, 2020; Hsu and Lin, 2023). This clarified that users are more satisfied with Chatbot when they receive their own questions, clear answers, trust the source, and have a personalized interaction. This would consequently make the user more willing to continue using chatbot services over time (Silva et al., 2023).

Furthermore, the results show that PAI mediates the relationship between AICSQ and chatbot CI. The findings indicate that very high service quality, accuracy, and personalized service of the chatbot are linked to users’ perceptions of the chatting robot's intelligence (Bernabei et al., 2023; Sohail et al., 2023), which in turn promotes positive trust and confidence, thereby influencing chatbot CI, while low service quality diminishes this perception (Chen et al., 2023). This mediating role of AI Intelligence represents a novel and important finding of the present study.

In addition, the results show that PT mediates the relationship between AICSQ and chatbot CI, which is consistent with prior studies (Ashfaq et al., 2020; Nguyen et al., 2021; Shahzad et al., 2024). This may be achieved by converting service attributes, such as accuracy, timeliness, reliability, assurance, and personalization, into users’ confidence in the chatbot's reliability, security, and overall performance (Nguyen et al., 2021; Trawnih et al., 2022). If it is perceived that chatbots are trustworthy, can provide up-to-date information, and can protect personal data, then their trust grows, which consequently strengthens their intention to keep using a chatbot. In contrast, poor service quality, misinformation, and security breaches erode trust, leading users to discontinue using a chatbot (Guo and Erdenebold, 2025; Kuberkar and Singhal, 2020; Rahim et al., 2022).

Furthermore, the study reveals that user satisfaction mediates the relationship between AICSQ and chatbot CI, consistent with prior research (Li et al., 2021; Silva and Canedo, 2024; Yu et al., 2024). User satisfaction connects the AICSQ and chatbot CI through system and service quality, including accuracy, reliability, and smooth performance. This is achieved by users having positive experiences with a given chatbot, which, in turn, reduces effort and frustration (Yu et al., 2024). Satisfaction increases their confidence and loyalty, provided there are fewer opportunities for their expectations to be betrayed or ignored. This is why it serves as the basis for a stronger chatbot CI (Silva et al., 2023).

The study ultimately found that PAI, PT, and user satisfaction intervene in sequence in the relationship between AICSQ and chatbot CI, a novel finding. High-quality chatbot services appreciably enhance PAI (Bernabei et al., 2023; Sohail et al., 2023), which, in turn, increases PT by decreasing uncertainty and strengthening confidence in reliability and privacy (Lu et al., 2025). Trust is the crucial factor for enhancing user satisfaction (Zhang et al., 2022), and when users are satisfied, they are more likely to continue using chatbot (Ngo et al., 2025).

6 Theoretical contributions and managerial implications

6.1 Theoretical contributions

This study contributes to the development of theory on the acceptance and CI of chatbots in several ways. First, this study is among the early efforts to examine actual chatbot users’ CI from the perspective of the S-O-R model. One of the strengths of the S-O-R model is its potential to illustrate cognitive states (AICSQ, PAI, PT) and affective states (satisfaction), as mediating variables. The S-O-R model goes beyond the technology acceptance model (TAM), which emphasizes perceived usefulness and perceived ease of use. The findings reveal that affective paths contribute more to chatbot CI than cognitive paths (Dužević et al., 2025). Second, the S-O-R model provides an effective bridge between the construct of the Information Systems Success Model (ISS) (particularly service quality) and the concept of behavioral intention (Alharbi et al., 2025). Third, previous studies on Chatbots have been more inclined towards literature reviews than towards empirical analysis (Maheshwari, 2024); however, this paper sheds light on factors through analysis of primary data. By examining students’ perspectives and behaviors, the paper has addressed the research gap identified in the literature. Moreover, the paper highlights the theoretical importance of various factors that contribute to the CI of Chatbots. Fourth, although studies have examined Chatbots in recent years, very few have examined the mechanisms by which AICSQ creates Chatbot CI among students. In particular, this study considers AICSQ as the stimulus, PAI, PT, and satisfaction as the organism, and CI as the response. The results show that the S-O-R model is a suitable framework for explaining users’ CI of AI chatbot beyond the traditional retail and consumer behavior context (Donovan and Rossiter, 2002; Eroglu et al., 2001; Kim et al., 2020). These findings also offer valuable insights into how generative AI can address the requirements of human-computer interaction, thereby making significant contributions to the existing literature (Yu et al., 2024). Fifth, this study focuses on capturing the service-based nature of AI Chatbots, their intelligence, PT, and satisfaction by using the S-O-R model combined with serial mediation to predict CI (Ashfaq et al., 2020; Bilquise et al., 2024; Kwangsawad and Jattamart, 2022; Nguyen et al., 2021; Pillai and Sivathanu, 2020; Rajaobelina et al., 2021; Rese et al., 2020; Silva et al., 2023). The study finds that AICSQ directly and indirectly impacts CI, underscoring its key importance. The current study reveals that PAI functions as both a direct determinant of CI and a mediator of CI. In this case, users’ perception of AI chatbot intelligence, including its ability to understand context, learn from interactions, and provide smart replies, strongly increases the likelihood that users will continue to intend to use the technology (Jiang et al., 2022). PAI can be considered a novel psychological mechanism distinguishing AI chatbot from other types of information systems. As the findings show, PT is another important determinant of CI and a mediator in the process. Thus, trusting in the Chatbot’ reliability, competence, and ethics significantly increases the chances of CI (Shin, 2021). Given the recent privacy issues related to AI, trust is a crucial mechanism for fostering the connection between AICSQ and CI. In the current study, satisfaction was the strongest predictor of continuance intention, confirming previous research findings in the context of ECM theory (Li et al., 2021). The final contribution of the study lies in the proof of sequential mediation: AICSQ→PAI→PT→Satisfaction→CI. Based on the sequence of effects, it can be concluded that AI chatbot service quality first influences PAI, creating the conditions for building trust, achieving satisfaction, and forming CI. In this regard, it should be noted that most traditional models ignore complex internal processes, which makes the proposed approach better at explaining technology continuance behavior. It should also be noted that this study provides contextual contribution through its validation among Saudi Arabian higher- education students in relation to Vision 2030 goals (Mohammed and Ferraris, 2025)

6.2 Managerial implications

A number of important managerial implications can be drawn from this study's findings. First, quality assurance specialists, software developers, and e-learning managers should focus on improving the quality of AI chatbots, information reliability, and pedagogical integration to enhance chatbot CI. chatbot systems such as ChatGPT, DeepSeek, Bashayer, Tayseer, and Nabiha should provide accurate, complete, timely, and context-aware responses through continuous knowledge base updates, prompt optimization, and feedback mechanisms, as service quality and information satisfaction strongly influence usability, trust, and chatbot CI (Alabbas and Alomar, 2024). Second, increasing PAI is essential for improving the meaningfulness of interaction. Chatbots should provide accurate, relevant, and context-aware responses, offering logical, practical solutions aligned with students’ academic needs. High levels of interactivity and intelligent responses not only make personalized learning possible but also help develop students’ confidence in chatbot capabilities, contributing to a positive learning experience. Third, intuitive design, structured interfaces, personalized interactions, and transparent data policies should be prioritized to improve trust and user satisfaction while reducing uncertainty and privacy concerns (Guo and Erdenebold, 2025; Okonkwo and Ade-Ibijola, 2021; Rafiq et al., 2022; Yu et al., 2024). Fourth, institutions should establish robust service recovery mechanisms, including error correction, escalation to human support, and balanced anthropomorphic features to manage user expectations and reduce dissatisfaction during service failures in AI chatbots such as ChatGPT and DeepSeek (Leong et al., 2026; Zhang et al. 2024b). For Saudi higher education, localized systems such as Tayseer should use Arabic NLP pipelines and culturally relevant content to improve usability and satisfaction among students.

Similarly, the Bashayer chatbot demonstrated that proper pedagogical integration improves students’ motivation and learning strategies (Al-Abdullatif et al., 2023; Alabbas and Alomar, 2024; Dužević et al., 2025). Finally, faculty and student training programs should promote the effective and ethical use of ChatGPT, DeepSeek, and other academic chatbots, and align chatbot tasks with course objectives and assessments. Providing faculty with sufficient pedagogical support is necessary, as inadequate training limits effective CI. Overall, high- quality, pedagogically integrated chatbot services improve PAI, thereby strengthening trust and user satisfaction, and sustaining chatbot use in higher education.

7 Limitations and future directions

Although it made contributions, this study has certain shortcomings that can guide future research. First, a key limitation is the use of convenience sampling and a small sample size. A limitation of convenience sampling is that participants are selected based on accessibility and willingness to participate rather than by random selection, which may introduce sampling bias and reduce the sample's representativeness. As a result, the findings may not be fully generalizable to the broader population (Andrade, 2021). Second limitation of the present study relates to the issue of common method variance because only one questionnaire survey was conducted at a point in time. Even though there was no evidence of common method variance in this study, it is important to note that the risk of such problem still exists. The use of other data collection methods could help avoid this problem in future studies (Kock et al., 2021).Third, the cross-sectional design captures students’ perceptions at a single point in time. It does not account for changes in familiarity, trust, or usage intentions as exposure to chatbot increases. Students with little prior experience might rely mainly on their first impressions or even preconceptions when forming their attitudes and intentions. Therefore, the findings capture only a glimpse of how students perceive the application at its early stages. Research employing longitudinal and experimental designs will help gain insight into why attitudes, trust, and continuance intentions evolve, as well as the validity of self- reported data. Fourth, cross-national, cross-group, or cross-sector analyses of chatbot CI may be pursued in the future (e.g., healthcare chatbot vs. tourism chatbot), which can help identify differences in chatbot continuance behavior. Finally, this study was limited to some factors affecting students’ chatbot acceptance, and no moderating variables were considered, which could affect the proposed relationships. To further improve its predictive power, future research should explore additional predictors, mediators, and moderators focused on privacy and misuse risk perceptions, with an emphasis on chatbot technology in its infancy.

8 Conclusion

This study sought to address a knowledge gap regarding the factors that influence students’ intention to continue using AI chatbots in higher education in the Kingdom of Saudi Arabia and to develop and validate an S-O-R–based model for this purpose. The results of this study provide evidence that AICSQ is a significant factor that directly and indirectly influences continuance intention. More specifically, high-quality chatbot services improve students’ perceptions of AI intelligence, build trust, and increase user satisfaction, creating a strong feedback loop that drives continuous use of chatbot systems. From a practical perspective, these findings offer actionable guidance for educators, developers, and institutions on designing, implementing, and managing chatbot systems that enhance student engagement and long-term CI. With approximately half of the variability in continuance intention accounted for by the model, there is strong empirical evidence that students would continue using chatbot if the systems were intelligent, reliable, trustworthy, and capable of delivering satisfactory learning experiences. The results indicate that the continuous use of AI chatbots in the higher education sector is not just about the technical side but also about confidence-building, satisfaction- guaranteeing, addressing emerging concerns, and privacy and risk considerations.

References

• 1

  AbdelkaderO. A. (2023). ChatGPT’s influence on customer experience in digital marketing: investigating the moderating roles. Heliyon9 (8), e18770. 10.1016/j.heliyon.2023.e18770

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  AcayıpE. (2024). Yapay Zeka Destekli Chatbot Hizmet Kalitesinin Müşteri Memnuniyetine Etkisi. Curr. Perspect. Soc. Sci.28 (4), 477–490. 10.53487/atasobed.1438079

  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  AcharyaA. S.PrakashA.SaxenaP.NigamA. (2013). Sampling: why and how of it. Indian J. Med. Spec.4 (2), 330–333.

  • Google Scholar
  • View reference in article

• 4

  AhnH. Y. (2025). Modeling student loyalty in the age of generative AI: a structural equation analysis of ChatGPT’s role in higher education. Systems.13 (10), 915. 10.3390/systems13100915

  • CrossRef
  • Google Scholar
  • View reference in article

• 5

  AkpanI. J.KobaraY. M.OwolabiJ.AkpanA. A.OffodileO. F. (2025). Conversational and generative artificial intelligence and human–chatbot interaction in education and research. Int. Trans. Oper. Res.32 (3), 1251–1281. 10.1111/itor.13522

  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  Al-AbdullatifA. M. (2023). Modeling students’ perceptions of chatbots in learning: integrating technology acceptance with the value-based adoption model. Educ. Sci.13 (11), 1151. 10.3390/educsci13111151

  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  Al-AbdullatifA. M.Al-DokhnyA. A.DrwishA. M. (2023). Implementing the Bashayer chatbot in Saudi higher education: measuring the influence on students’ motivation and learning strategies. Front. Psychol.14, 1129070. 10.3389/fpsyg.2023.1129070

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  Al-AmriN. A.Al-AbdullatifA. M. (2024). Drivers of chatbot adoption among K–12 teachers in Saudi Arabia. Educ. Sci.14 (9), 1034. 10.3390/educsci14091034

  • CrossRef
  • Google Scholar
  • View reference in article

• 9

  Al-SharafiM. A.Al-EmranM.IranmaneshM.Al-QaysiN.IahadN. A.ArpaciI. (2023). Understanding the impact of knowledge management factors on the sustainable use of AI-based chatbots for educational purposes using a hybrid SEM-ANN approach. Interact. Learn. Environ.31 (10), 7491–7510. 10.1080/10494820.2022.2075014

  • CrossRef
  • Google Scholar
  • View reference in article

• 10

  Al-TahitahA. N.Al-SharafiM. A.AbdulrabM. (2021). “How COVID-19 pandemic is accelerating the transformation of higher education institutes: a health belief model view,” in Emerging Technologies During the Era of COVID-19 Pandemic, (Cham: Springer), 333–347.

  • Google Scholar
  • View reference in article

• 11

  AlabbasA.AlomarK. (2024). Tayseer: a novel AI-powered Arabic chatbot framework for technical and vocational student helpdesk services and enhancing student interactions. Appl. Sci.14 (6), 2547. 10.3390/app14062547

  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  AlagarsamyS.MehroliaS. (2023). Exploring chatbot trust: antecedents and behavioural outcomes. Heliyon9 (5), e16074. 10.1016/j.heliyon.2023.e16074

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 13

  AlharbiK.KhalilL.IslamM.MleikiA. (2025). Cultural influence on ChatGPT effectiveness in higher education: a Pakistan-Saudi Arabia comparison. Forum Linguist. Stud.7 (10), 372–389. 10.30564/fls.v7i10.10356

  • CrossRef
  • Google Scholar
  • View reference in article

• 14

  AlkadiR. S.AbedS. S. (2025). Examining consumer intention to adopt AI-powered chatbots in the Saudi banking sector: the moderating role of knowledge in technology. Int. J. Bank Mark.43 (10), 2229–2255. 10.1108/IJBM-10-2024-0635

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  AlmufarrehA. (2024). Determinants of students’ satisfaction with AI tools in education: a PLS-SEM-ANN approach. Sustainability.16 (13), 5354. 10.3390/su16135354

  • CrossRef
  • Google Scholar
  • View reference in article

• 16

  AlotaibiH. M.SonbulS. S.El-DakhsD. A. (2025). Factors influencing the acceptance and use of ChatGPT among English as a foreign language learners in Saudi Arabia. Humanit. Soc. Sci. Commun.12 (1), 1–13. 10.1057/s41599-025-04945-2

  • CrossRef
  • Google Scholar
  • View reference in article

• 17

  AlzahraniS.BhuniaA. (2025a). Fintech adoption intention among Gen Z in Saudi Arabia: examining the serial mediation of user innovativeness, perceived ease of use, trust, and usefulness. Qubahan Acad. J.5 (3), 559–579. 10.48161/qaj.v5n3a1979

  • CrossRef
  • Google Scholar
  • View reference in article

• 18

  AlzahraniS.BhuniaA. K. (2024). A serial mediation model of the relationship between digital entrepreneurial education, alertness, motivation, and intentions. Sustainability.16 (20), 8858. 10.3390/su16208858

  • CrossRef
  • Google Scholar
  • View reference in article

• 19

  AlzahraniS.BhuniaA. K. (2025b). An integrated model of Fintech adoption: examining the dual serial mediation of digital literacy, ease of use, usefulness, and perceived value among Gen Z students in Saudi Arabia. Educ. Process: Int. J.17, e2025370.

  • Google Scholar
  • View reference in article

• 20

  AminM. A.KimY. S.NohM. (2025). Unveiling the drivers of ChatGPT utilization in higher education sectors: the direct role of perceived knowledge and the mediating role of trust in ChatGPT. Educ. Inf. Technol.30 (6), 7265–7291. 10.1007/s10639-024-13095-y

  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  AmorosoD.LimR. (2017). The mediating effects of habit on continuance intention. Int. J. Inf. Manage.37 (6), 693–702. 10.1016/j.ijinfomgt.2017.05.003

  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  AndersonJ. C.GerbingD. W. (1988). Structural equation modeling in practice: a review and recommended two-step approach. Psychol. Bull.103 (3), 411–423. 10.1037/0033-2909.103.3.411

  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  AndradeC. (2021). The inconvenient truth about convenience and purposive samples. Indian. J. Psychol. Med.43 (1), 86–88. 10.1177/0253717620977000PMID: 34349313 PMCID: PMC8295573.

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 24

  AnwarI.AhmadA.SaleemI.YasinN. (2023). Role of entrepreneurship education, passion and motivation in augmenting omani students’ entrepreneurial intention: a stimulus-organism-response approach. Int. J. Manag. Educ.21 (3), 100842. 10.1016/j.ijme.2023.100842

  • CrossRef
  • Google Scholar
  • View reference in article

• 25

  AraújoT.CasaisB. (2019). “Customer acceptance of shopping-assistant chatbots,” in Marketing and Smart Technologies: Proceedings of ICMarkTech 2019, (Singapore: Springer), 278–287.

  • Google Scholar
  • View reference in article

• 26

  AshfaqM.YunJ.YuS.LoureiroS. M. C. (2020). I, chatbot: modeling the determinants of users’ satisfaction and continuance intention of AI-powered service agents. Telemat. Inform.54, 101473. 10.1016/j.tele.2020.101473

  • CrossRef
  • Google Scholar
  • View reference in article

• 27

  AskJ. A.FacemireM.HoganA.ConversationsH. B. (2016). The state of chatbots. Forrester. Com Report20, 1–16.

  • Google Scholar
  • View reference in article

• 28

  AsquerA.KrachkovskayaI. (2022). Designing public financial management systems: exploring the use of chatbot-assisted case studies. Public Money Manag.42 (7), 551–557. 10.1080/09540962.2022.2069412

  • CrossRef
  • Google Scholar
  • View reference in article

• 29

  AthotaL.ShuklaV. K.PandeyN.RanaA. (2020). “Chatbot for healthcare system using artificial intelligence,” in 2020 8th International Conference on Reliability, Infocom Technologies and Optimization (Trends and Future Directions) (ICRITO), (Noida: IEEE), 619–622.

  • Google Scholar
  • View reference in article

• 30

  AvcılarM. Y.YenilmezG. (2026). The effects of chatbot characteristics on satisfaction and continuance intention: the moderating role of the need for human interaction. J. Theor. Appl. Electron. Commer. Res.21 (4), 122. 10.3390/jtaer21040122

  • CrossRef
  • Google Scholar
  • View reference in article

• 31

  AyanwaleM. A. (2024). “Using diffusion theory of innovation to investigate perceptions of STEM and non-STEM students,” in 2024 IEEE Global Engineering Education Conference (EDUCON), Kos Island, Greece, 2024, 1–8. 10.1109/EDUCON60312.2024.10578835

  • CrossRef
  • Google Scholar
  • View reference in article

• 32

  BagozziR. P.YiY. (1988). Onthe evaluation of structural equation models. J. Acad. Mark. Sci.16 (1), 74–94. 10.1007/BF02723327

  • CrossRef
  • Google Scholar
  • View reference in article

• 33

  BalaskasS.TsiantosV.ChatzifotiouS.RigouM. (2025). Determinants of ChatGPT adoption intention in higher education: expanding on TAM with the mediating roles of trust and risk. Information.16 (2), 82. 10.3390/info16020082

  • CrossRef
  • Google Scholar
  • View reference in article

• 34

  BaroniI.CalegariG. R.ScandolariD.CelinoI. (2022). AI-TAM: a model to investigate user acceptance and collaborative intention in human-in-the- loop AI applications. Hum. Comput.9 (1), 1–21. 10.15346/hc.v9i1.134

  • CrossRef
  • Google Scholar
  • View reference in article

• 35

  BaroudiJ. J.OrlikowskiW. J. (1989). The problem of statistical power in MIS research. MIS. Q.13, 87–106. 10.2307/248704

  • CrossRef
  • Google Scholar
  • View reference in article

• 36

  BartneckC.KulićD.CroftE.ZoghbiS. (2009). Measurement instruments for the anthropomorphism, animacy, likeability, perceived intelligence, and perceived safety of robots. Int. J. Soc. Robot.1 (1), 71–81. 10.1007/s12369-008-0001-3

  • CrossRef
  • Google Scholar
  • View reference in article

• 37

  BenitezJ.HenselerJ.CastilloA.SchuberthF. (2020). How to perform and report an impactful analysis using partial least squares: guidelines for confirmatory and explanatory IS research. Inf. Manag.57 (2), 103168. 10.1016/j.im.2019.05.003

  • CrossRef
  • Google Scholar
  • View reference in article

• 38

  BernabeiM.ColabianchiS.FalegnamiA.CostantinoF. (2023). Students’ use of large language models in engineering education: a case study on technology acceptance, perceptions, efficacy, and detection chances. Comput. Educ.: Artif. Intell.5, 100172. 10.1016/j.caeai.2023.100172

  • CrossRef
  • Google Scholar
  • View reference in article

• 39

  BhattacherjeeA. (2001). Understanding information systems continuance: an expectation-confirmation model. MIS. Q.25 (3), 351–370. 10.2307/3250921

  • CrossRef
  • Google Scholar
  • View reference in article

• 40

  BhuniaA. K.ShomeM. K. (2023). Anempirical comparative study between the theory of planned behavior and psychological capital in predicting entrepreneurial intention. J. Syst. Manag. Sci.14 (4), 447–468.

  • Google Scholar
  • View reference in article

• 41

  BilquiseG.IbrahimS.SalhiehS. M. (2024). Investigating student acceptance of an academic advising chatbot in higher education institutions. Educ. Inf. Technol.29 (5), 6357–6382. 10.1007/s10639-023-12076-x

  • CrossRef
  • Google Scholar

• 42

  BiswasA.VermaR. K. (2021). Attitude and alertness in personality traits: a pathway to building entrepreneurial intentions among university students. J. Entrep.30 (2), 367–396. 10.1177/09713557211025656

  • CrossRef
  • Google Scholar

• 43

  BoomsmaA. (1983). On the robustness Of LISREL (maximum likelihood estimation) against small sample size and non-normality. (Unpublished Ph.D. dissertation). University of Groningen, Groningen.

  • Google Scholar

• 44

  ByrneB. M. (2013). Structural Equation Modelling with Mplus: Basic Concepts, Applications, and Programming. 2nd ed. New York: Routledge.

  • Google Scholar

• 45

  CalvaresiD.IbrahimA.CalbimonteJ.-P.FragniereE.ScheggR.SchumacherM. I. (2023). Leveraging inter-tourists interactions via chatbots to bridge academia, tourism industries and future societies. J. Tour. Futures.9 (3), 311–337. 10.1108/JTF-01-2021-0009

  • CrossRef
  • Google Scholar

• 46

  ChawL. Y.TangC. M. (2018). What makes learning management systems effective for learning?J. Educ. Technol. Syst.47 (2), 152–169. 10.1177/0047239518795828

  • CrossRef
  • Google Scholar

• 47

  ChenQ.GongY.LuY.TangJ. (2022). Classifying and measuring the service quality of AI chatbot in frontline service. J. Bus. Res.145, 552–568. 10.1016/j.jbusres.2022.02.088

  • CrossRef
  • Google Scholar

• 48

  ChenQ.LuY.GongY.XiongJ. (2023). Can AI chatbots help retain customers? Impact of AI service quality on customer loyalty. Internet Res.33 (6), 2205–2243. 10.1108/INTR-09-2021-0686

  • CrossRef
  • Google Scholar

• 49

  ChengJ. W.MitomoH. (2017). The underlying factors of the perceived usefulness of using smart wearable devices for disaster applications. Telemat. Inform.34 (2), 528–539. 10.1016/j.tele.2016.09.010

  • CrossRef
  • Google Scholar

• 50

  ChengY.JiangH. (2020). How do AI-driven chatbots impact user experience? Examining gratifications, perceived privacy risk, satisfaction, loyalty, and continued use. J. Broadcast. Electron. Media.64 (4), 592–614. 10.1080/08838151.2020.1834296

  • CrossRef
  • Google Scholar

• 51

  ChinW. W.GopalA.SalisburyW. D. (1997). Advancing the theory of adaptive structuration: the development of a scale to measure faithfulness of appropriation. Inf. Syst. Res.8 (4), 342–367. 10.1287/isre.8.4.342

  • CrossRef
  • Google Scholar

• 52

  ChoiY.-s.LeeS.-z.ChoiJ. (2025). A study on factors influencing continuous usage intention of chatbot services in South Korean financial institutions. Int. J. Financ. Stud.13 (2), 56. 10.3390/ijfs13020056

  • CrossRef
  • Google Scholar

• 53

  ChopdarP. K.BalakrishnanJ. (2020). Consumers response towards Mobile commerce applications: SOR approach. Int. J. Inf. Manage.53, 102106. 10.1016/j.ijinfomgt.2020.102106

  • CrossRef
  • Google Scholar

• 54

  ChotisarnN.PhuthongT. (2025). Impact of artificial intelligence- enabled service attributes on customer satisfaction and loyalty in chain hotels: evidence from coastal tourism destinations in Western Thailand. Soc. Sci. Humanit. Open.11, 101306. 10.1016/j.ssaho.2025.101306

  • CrossRef
  • Google Scholar

• 55

  ChoudhuryA.ShamszareH. (2023). Investigating the impact of user trust on the adoption and use of ChatGPT: survey analysis. J. Med. Internet. Res.25, e47184. 10.2196/47184

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 56

  ChungK.ParkR. C. (2019). Chatbot-based heathcare service with a knowledge base for cloud computing. Cluster. Comput.22 (1), 1925–1937. 10.1007/s10586-018-2334-5

  • CrossRef
  • Google Scholar

• 57

  ChungM.KoE.JoungH.KimS. J. (2020). Chatbot e-service and customer satisfaction regarding luxury brands. J. Bus. Res.117, 587–595. 10.1016/j.jbusres.2018.10.004

  • CrossRef
  • Google Scholar

• 58

  CohenJ. (2013). Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Mahwah, NJ: Lawrence Erlbaum.

  • Google Scholar

• 59

  CohenJ.CohenP.WestS. G.AikenL. S. (2013). Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences. 3rd ed. Mahwah, NJ: L. Erlbaum Associates.

  • Google Scholar

• 60

  ÇeraG.NdokaM.DikaI.ÇeraE. (2022). Examining the impact of COVID-19 on entrepreneurial intention through a stimulus–organism–response perspective. Adm. Sci.12 (4), 184. 10.3390/admsci12040184

  • CrossRef
  • Google Scholar

• 61

  DahriN. A.YahayaN.Al-RahmiW. M.et al (2024). Extended TAM based acceptance of AI-powered ChatGPT for supporting metacognitive self- regulated learning in education: a mixed-methods study. Heliyon.10 (8), e29317. 10.1016/j.heliyon.2024.e29317

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 62

  DevisaktiA.MuftahuM. (2023). Digitalization in higher education: does personal innovativeness matter in digital learning?Interact. Technol. Smart Educ.20 (2), 257–270. 10.1108/ITSE-10-2021-0182

  • CrossRef
  • Google Scholar

• 63

  DingY.NajafM. (2024). Interactivity, humanness, and trust: a psychological approach to AI chatbot adoption in e-commerce. BMC. Psychol.12 (1), 595. 10.1186/s40359-024-02083-z

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 64

  DonovanR. J.RossiterJ. R. (2002). “Store atmosphere: an environmental psychology,” in Retailing: Critical Concepts 3.2. Retail Practices and Operations (Vol 2), 77.

  • Google Scholar

• 65

  DuževićI.BakovićT.SurmanV. (2025). Understanding artificial intelligence chatbot quality and experience: a higher education student perspective. Entrep. Bus. Econ. Rev.13 (3), 151–170. 10.15678/EBER.2025.130308

  • CrossRef
  • Google Scholar

• 66

  ElayatA. M. A.ElalfyR. M. (2025). Using SOR theory to examine the impact of AI chatbot quality on GenZ’s satisfaction and advocacy within the fast-food sector. Young Consum.26 (2), 352–383. 10.1108/YC-08-2024-2199

  • CrossRef
  • Google Scholar

• 67

  ElnadiM.GheithM. H. (2023). The role of individual characteristics in shaping digital entrepreneurial intention among university students: evidence from Saudi Arabia. Think. Ski. Creat.47, 101236. 10.1016/j.tsc.2023.101236

  • CrossRef
  • Google Scholar

• 68

  ErogluS. A.MachleitK. A.DavisL. M. (2001). Atmospheric qualities of online retailing: a conceptual model and implications. J. Bus. Res.54 (2), 177–184. 10.1016/S0148-2963(99)00087-9

  • CrossRef
  • Google Scholar

• 69

  FangY.QureshiI.SunH.McColeP.RamseyE.LimK. H. (2014). Trust, satisfaction, and online repurchase intention. MIS. Q.38 (2), 407–438. 10.25300/MISQ/2014/38.2.04

  • CrossRef
  • Google Scholar

• 70

  FengZ.HuG.LiB.WangJ. (2025). Unleashing the power of ChatGPT in finance research: opportunities and challenges. Financ. Innov.11 (1), 93. 10.1186/s40854-025-00770-3

  • CrossRef
  • Google Scholar

• 71

  FerreiraM.BarbosaB. (2023). “A review on chatbot personality and its expected effects on users,” in Trends, Applications, and Challenges of Chatbot Technology, 222–243. 10.4018/978-1-6684-6234-8.ch010

  • CrossRef
  • Google Scholar

• 72

  FornellC.LarckerD. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. J. Mark. Res.18 (1), 39–50. 10.1177/002224378101800104

  • CrossRef
  • Google Scholar

• 73

  ForoughiB.IranmaneshM.GhobakhlooM.SenaliM. G.AnnamalaiN.Naghmeh-AbbaspourB.et al (2025). Determinants of ChatGPT adoption among students in higher education: the moderating effect of trust. Electron. Libr.43 (1), 1–21. 10.1108/EL-12-2023-0293

  • CrossRef
  • Google Scholar

• 74

  Fui-Hoon NahF.ZhengR.CaiJ.SiauK.ChenL. (2023). Generative AI and ChatGPT: applications, challenges, and AI-human collaboration. J. Inf. Technol. Case Appl. Res.25 (3), 277–304. 10.1080/15228053.2023.2233814

  • CrossRef
  • Google Scholar

• 75

  GatzioufaP.SaprikisV. (2022). A literature review on users’ behavioral intention toward chatbots’ adoption. Appl. Comput. Inform. (ahead of print). 10.1108/ACI-01-2022-0021

  • CrossRef
  • Google Scholar

• 76

  GefenD.KarahannaE.StraubD. W. (2003). Inexperience and experience with online stores: the importance of TAM and trust. IEEE Trans. Eng. Manag.50 (3), 307–321. 10.1109/TEM.2003.817277

  • CrossRef
  • Google Scholar

• 77

  GefenD.StraubD.BoudreauM.-C. (2000). Structural equation modeling and regression: guidelines for research practice. Commun. Assoc. Inf. Syst.4 (1), 1–77. 10.17705/1CAIS.00407

  • CrossRef
  • Google Scholar

• 78

  GrahamG.NisarT. M.PrabhakarG.MeritonR.MalikS. (2025). Chatbots in customer service within banking and finance: do chatbots herald the start of an AI revolution in the corporate world?Comput. Human. Behav.165, 108570. 10.1016/j.chb.2025.108570

  • CrossRef
  • Google Scholar

• 79

  GuoH.ErdeneboldT. (2025). Factors influencing intention to adopt an AI chatbot for learning in higher education: an integrated PLS-SEM, IPMA, and ANN approach. Comput. Educ. Artif. Intell.9, 100477. 10.1016/j.caeai.2025.100477

  • CrossRef
  • Google Scholar

• 80

  HairJ. F.BlackW. C.BabinB. J.AndersonR. E. (2019). Multivariate Data Analysis. 8th ed. Boston: Cengage.

  • Google Scholar

• 81

  HanT.KimD.AhnJ.RyuK.LeeS. (2025). A study on the effect of voicebot quality on intention to continuous use in AI contact centers in the financial sector. KSII Trans. Internet Inf. Syst.19 (3), 1027. 10.3837/tiis.2025.03.017

  • CrossRef
  • Google Scholar

• 82

  HarmanH. H. (1976). Modern Factor Analysis. Chicago, IL: University of Chicago Press.

  • Google Scholar

• 83

  HassaniH.SilvaE. S. (2023). The role of ChatGPT in data science: how ai-assisted conversational interfaces are revolutionizing the field. Big Data Cogn. Comput.7 (2), 62. 10.3390/bdcc7020062

  • CrossRef
  • Google Scholar

• 84

  HenselerJ.RingleC. M.SarstedtM. (2015). A new criterion for assessing discriminant validity in variance-based structural equation modeling. J. Acad. Mark. Sci.43 (1), 115–135. 10.1007/s11747-014-0403-8

  • CrossRef
  • Google Scholar

• 85

  HenselerJ.RingleC. M.SinkovicsR. R. (2009). “The use of partial least squares path modeling in international marketing,” in New Challenges to International Marketing, eds. SinkovicsR. R.GhauriP. N. (Emerald Group Publishing Limited), 277–319.

  • Google Scholar

• 86

  HsiaoK.-L.ChenC.-C. (2022). What drives continuance intention to use a food-ordering chatbot? An examination of trust and satisfaction. Libr. Hi Tech.40 (4), 929–946. 10.1108/LHT-08-2021-0274

  • CrossRef
  • Google Scholar

• 87

  HsuC.-L.LinJ. C.-C. (2023). Understanding the user satisfaction and loyalty of customer service chatbots. J. Retail. Consum. Serv.71, 103211. 10.1016/j.jretconser.2022.103211

  • CrossRef
  • Google Scholar

• 88

  HuL.BentlerP. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct. Equ. Model. Multidiscip. J.6 (1), 1–55. 10.1080/10705519909540118

  • CrossRef
  • Google Scholar

• 89

  HuangD.-H.ChuehH.-E. (2021). Chatbot usage intention analysis: veterinary consultation. J. Innov. Knowl.6 (3), 135–144. 10.1016/j.jik.2020.09.002

  • CrossRef
  • Google Scholar

• 90

  HuangS. Y. (2025). Why can Fintech chatbots guide consumers to buy banking products?Int. J. Hum.–Comput. Interact., 41 (19), 12229–12234.

  • Google Scholar

• 91

  HuangS. Y. B.LeeC.-J. (2022). Predicting continuance intention to Fintech chatbot. Comput. Human. Behav.129, 107027. 10.1016/j.chb.2021.107027

  • CrossRef
  • Google Scholar

• 92

  JangY.-T. J.LiuA. Y.KeW.-Y. (2023). Exploring smart retailing: anthropomorphism in voice shopping of smart speaker. Inf. Technol. People.36 (7), 2894–2913. 10.1108/ITP-07-2021-0536

  • CrossRef
  • Google Scholar

• 93

  JeonJ. (2024). Exploring AI chatbot affordances in the EFL classroom: young learners’ experiences and perspectives. Comput. Assist. Lang. Learn.37 (1–2), 1–26. 10.1080/09588221.2021.2021241

  • CrossRef
  • Google Scholar

• 94

  JiangH.ChengY.YangJ.GaoS. (2022). AI-powered chatbot communication with customers: dialogic interactions, satisfaction, engagement, and customer behavior. Comput. Human. Behav.134, 107329. 10.1016/j.chb.2022.107329

  • CrossRef
  • Google Scholar

• 95

  JohnsonK.LiY.PhanH.SingerJ.TrinhH. (2012). The innovative success that Is Apple, Inc.

  • Google Scholar

• 96

  JuitaR.InanD. I.SantosoB. (2026). Digital market adoption by underserved MSMEs in developing countries: mediation and moderation by self-efficacy and trust. Int. J. Inf. Manag. Data Insights.6 (1), 100384. 10.1016/j.jjimei.2025.100384

  • CrossRef
  • Google Scholar

• 97

  JyothsnaM.Subbaiah PV.KryvinskaN. (2024). Exploring the chatbot usage intention-a mediating role of chatbot initial trust. Heliyon.10 (12), e33028. 10.1016/j.heliyon.2024.e33028

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 98

  KaabachiS.Ben MradS.O’LearyB. (2019). Consumer’s initial trust formation in IOB’s acceptance: the role of social influence and perceived compatibility. Int. J. Bank Mark.37 (2), 507–530. 10.1108/IJBM-12-2017-0270

  • CrossRef
  • Google Scholar

• 99

  KasilingamD. L. (2020). Understanding the attitude and intention to use smartphone chatbots for shopping. Technol. Soc.62, 101280. 10.1016/j.techsoc.2020.101280

  • CrossRef
  • Google Scholar

• 100

  KavithaK.JoshithV. P. (2025). Exploring GenZ’s satisfaction and continuance intention towards ChatGPT 4o in higher education: PLS-SEM insights using the ECM. J. Educ. Technol. Syst.54 (2), 407–430. 10.1177/00472395251380294

  • CrossRef
  • Google Scholar

• 101

  KhanI. U.HameedZ.YuY.KhanS. U. (2017). Assessing the determinants of flow experience in the adoption of learning management systems: the moderating role of perceived institutional support. Behav. Inf. Technol.36 (11), 1162–1176. 10.1080/0144929X.2017.1362475

  • CrossRef
  • Google Scholar

• 102

  KimB. (2019). Understanding key antecedents of consumer loyalty toward sharing-economy platforms: the case of airbnb. Sustainability.11 (19), 5195. 10.3390/su11195195

  • CrossRef
  • Google Scholar

• 103

  KimJ.LennonS. J. (2013). Effects of reputation and website quality on online consumers’ emotion, perceived risk and purchase intention: based on the stimulus-organism-response model. J. Res. Interact. Mark.7 (1), 33–56. 10.1108/17505931311316734

  • CrossRef
  • Google Scholar

• 104

  KimM.-J.ChungN.LeeC.-K. (2011). The effect of perceived trust on electronic commerce: shopping online for tourism products and services in South Korea. Tour. Manag.32 (2), 256–265. 10.1016/j.tourman.2010.01.011

  • CrossRef
  • Google Scholar

• 105

  KimM. J.LeeC.-K.JungT. (2020). Exploring consumer behavior in virtual reality tourism using an extended stimulus-organism-response model. J. Travel Res.59 (1), 69–89. 10.1177/0047287518818915

  • CrossRef
  • Google Scholar

• 106

  KimS. Y.KimJ. (2025). The impact of AI recommendation quality on service satisfaction: the moderating roles of standardization and customization. J. Serv. Mark.39 (4), 365–386. 10.1108/JSM-05-2024-0214

  • CrossRef
  • Google Scholar

• 107

  KlineR. B. (2023). Principles and Practice of Structural Equation Modeling. (2nd ed.). New York, NY: Guilford publications.

  • Google Scholar

• 108

  KockF.BerbekovaA.AssafA. G. (2021). Understanding and managing the threat of common method bias: detection, prevention, and control. Tour. Manag.86, 104330. 10.1016/j.tourman.2021.104330

  • CrossRef
  • Google Scholar

• 109

  KockN.HadayaP. (2018). Minimum sample size estimation in PLS-SEM: the inverse square root and gamma-exponential methods. Inf. Syst. J.28 (1), 227–261. 10.1111/isj.12131

  • CrossRef
  • Google Scholar

• 110

  KooliC. (2023). Chatbots in education and research: a critical examination of ethical implications and solutions. Sustainability.15 (7), 5614. 10.3390/su15075614

  • CrossRef
  • Google Scholar

• 111

  KuberkarS.SinghalT. K. (2020). Factors influencing adoption intention of AI powered chatbot for public transport services within a smart city. Int. J. Emerg. Technol. Learn.11 (3), 948–958.

  • Google Scholar

• 112

  KuhailM. A.FarooqS. A.AlmutairiS. (2023). Recent developments in chatbot usability and design methodologies. In Trends, Applications, and Challenges of Chatbot Technology, eds. KuhailM., ShawarB. A.HammadR.. (Hershey, PA: IGI Global), 1–23.

  • Google Scholar

• 113

  KwangsawadA.JattamartA. (2022). Overcoming customer innovation resistance to the sustainable adoption of chatbot services: a community-enterprise perspective in Thailand. J. Innov. Knowl.7 (3), 100211. 10.1016/j.jik.2022.100211

  • CrossRef
  • Google Scholar

• 114

  LaRoseR. (2010). The problem of media habits. Commun. Theory.20 (2), 194–222. 10.1111/j.1468-2885.2010.01360.x

  • CrossRef
  • Google Scholar

• 115

  LeeJ.-C.TangY.JiangS. (2023). Understanding continuance intention of artificial intelligence (AI)-enabled Mobile banking applications: an extension of AI characteristics to an expectation confirmation model. Humanit. Soc. Sci. Commun.10 (1), 1–12. 10.1057/s41599-023-01845-1

  • CrossRef
  • Google Scholar

• 116

  LeggS.HutterM. (2007). A collection of definitions of intelligence. Front. Artif. Intell. Appl.157, 17.

  • Google Scholar

• 117

  LeonardL. N.JonesK. (2021). Trust in C2C electronic commerce: ten years later. J. Comput. Inf. Syst.61 (3), 240–246. 10.1080/08874417.2019.1598829

  • CrossRef
  • Google Scholar

• 118

  LeongM. K.SidhuS. K.KhooK. L. (2026). AI chatbot service recovery quality and customer experience: a moderated mediation model of continuance usage intention. J. Consum. Mark.43 (2), 195–208. 10.1108/JCM-02-2025-7638

  • CrossRef
  • Google Scholar

• 119

  LiC.-Y.ZhangJ.-T. (2023). Chatbots or me? Consumers’ switching between human agents and conversational agents. J. Retail. Consum. Serv.72, 103264. 10.1016/j.jretconser.2023.103264

  • CrossRef
  • Google Scholar

• 120

  LiL.LeeK. Y.EmokpaeE.YangS.-B. (2021). What makes you continuously use chatbot services? Evidence from Chinese online travel agencies. Electron. Mark.31 (3), 575–599. 10.1007/s12525-020-00454-z

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 121

  LiW.BhuttoT. A.XuhuiW.MaitloQ.ZafarA. U.BhuttoN. A. (2020). Unlocking Employees’ green creativity: the effects of green transformational leadership, green intrinsic, and extrinsic motivation. J. Cleaner Prod.255, 120229. 10.1016/j.jclepro.2020.120229

  • CrossRef
  • Google Scholar

• 122

  LiY.LiY.ChenQ.ChangY. (2024). Humans as teammates: the signal of human–AI teaming enhances consumer acceptance of chatbots. Int. J. Inf. Manage.76, 102771. 10.1016/j.ijinfomgt.2024.102771

  • CrossRef
  • Google Scholar

• 123

  LimW. M.GunasekaraA.PallantJ. L.PallantJ. I.PechenkinaE. (2023). Generative AI and the future of education: ragnarök or reformation? A paradoxical perspective from management educators. Int. J. Manag. Educ.21 (2), 100790. 10.1016/j.ijme.2023.100790

  • CrossRef
  • Google Scholar

• 124

  LittleT. D.CardN. A.BovairdJ. A.PreacherK. J.CrandallC. S. (2007). “Structural equation modeling of mediation and moderation with contextual factors.” in Modeling Contextual Effects in Longitudinal Studies, 207–230.

  • Google Scholar

• 125

  LuC.-C. A.YehC.-C. R.LaiC.-C. S. (2025). The role of intelligence, trust and interpersonal job characteristics in employees’ AI usage acceptance. Int. J. Hosp. Manag.126, 104032. 10.1016/j.ijhm.2024.104032

  • CrossRef
  • Google Scholar

• 126

  LuZ.MinQ.JiangL.ChenQ. (2024). The effect of the anthropomorphic design of chatbots on customer switching intention when the chatbot service fails: an expectation perspective. Int. J. Inf. Manage.76, 102767. 10.1016/j.ijinfomgt.2024.102767

  • CrossRef
  • Google Scholar

• 127

  LuoC.HasanN. A. M.AhmadA. M. Z. (2024). Exploring satisfaction and trust as key drivers of e-government continuance intention: evidence from China for sustainable digital governance. Sustainability. 16 (24), 11068. 10.3390/su162411068

  • CrossRef
  • Google Scholar

• 128

  LynchS. M. (2003). Cohort and life-course patterns in the relationship between education and health: a hierarchical approach. Demography.40 (2), 309–331. 10.1353/dem.2003.0016

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 129

  MacCallumR. C.BrowneM. W.SugawaraH. M. (1996). Power analysis and determination of sample size for covariance structure modeling. Psychol. Methods.1 (2), 130. 10.1037/1082-989X.1.2.130

  • CrossRef
  • Google Scholar

• 130

  MacKenzieS. B.PodsakoffP. M. (2012). Common method bias in marketing: causes, mechanisms, and procedural remedies. J. Retail.88 (4), 542–555. 10.1016/j.jretai.2012.08.001

  • CrossRef
  • Google Scholar

• 131

  MacKinnonD. P.LockwoodC. M.WilliamsJ. (2004). Confidence limits for the indirect effect: distribution of the product and resampling methods. Multivariate. Behav. Res.39 (1), 99–128. 10.1207/s15327906mbr3901_4

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 132

  MageiraK.PittouD.PapasalourosA.KotisK.ZangogianniP.DaradoumisA. (2022). Educational AI chatbots for content and language integrated learning. Appl. Sci.12 (7), 3239. 10.3390/app12073239

  • CrossRef
  • Google Scholar

• 133

  MagnoF.DossenaG. (2023). The effects of chatbots’ attributes on customer relationships with brands: PLS-SEM and importance–performance map analysis. TQM J.35 (5), 1156–1169. 10.1108/TQM-02-2022-0080

  • CrossRef
  • Google Scholar

• 134

  MaheshwariG. (2024). Factors influencing students’ intention to adopt and use ChatGPT in higher education: a study in the Vietnamese context. Educ. Inf. Technol.29 (10), 12167–12195. 10.1007/s10639-023-12333-z

  • CrossRef
  • Google Scholar

• 135

  MalmströmH.StöhrC.OuW. (2023). Chatbots and other AI for learning: a survey of use and views among university students in Sweden. 10.17196/cls.csclhe/2023/01

  • CrossRef
  • Google Scholar

• 136

  ManigandanL.AlurS. (2024). An in-depth investigation into the influence of chatbot usability and age on continuous intention to use: a comprehensive study. Asia Pac. J. Inf. Syst.34 (1), 351–371. 10.14329/apjis.2024.34.1.351

  • CrossRef
  • Google Scholar

• 137

  McKnightD. H.ChoudhuryV.KacmarC. (2002). The impact of initial consumer trust on intentions to transact with a web site: a trust building model. J. Strateg. Inf. Syst.11 (3–4), 297–323. 10.1016/S0963-8687(02)00020-3

  • CrossRef
  • Google Scholar

• 138

  McTearM.CallejasZ.GriolD. (2016). “Conversational interfaces: devices, wearables, virtual agents, and robots,” in The Conversational Interface: Talking to Smart Devices, (Cham: Springer), 283–308.

  • Google Scholar

• 139

  MehrabianA.RussellJ. A. (1974). An Approach to Environmental Psychology. The MIT Press.

  • Google Scholar

• 140

  MehroliaS.AlagarsamyS.MoorthyV.JeevanandaS. (2023). Will users continue using banking chatbots? The moderating role of perceived risk. FIIB Bus Rev., 23197145231169900. 10.1177/23197145231169900

  • CrossRef
  • Google Scholar

• 141

  MehroliaS.AlagarsamyS.SabariM. I. (2021). Moderating effects of academic involvement in web-based learning management system success: a multigroup analysis. Heliyon.7 (5), e07000. 10.1016/j.heliyon.2021.e07000

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 142

  MimounM. S. B.PoncinI.GarnierM. (2017). Animated conversational agents and e-consumer productivity: the roles of agents and individual characteristics. Inf. Manag.54 (5), 545–559. 10.1016/j.im.2016.11.008

  • CrossRef
  • Google Scholar

• 143

  MohammedA.FerrarisA. (2025). Exploring motivational drivers of AI chatbot adoption in emerging markets: insights from the stimulus-organism- response model for service automation. Bottom Line.39 (3), 299–319. 10.1108/BL-11-2024-0189

  • CrossRef
  • Google Scholar

• 144

  MohammedI. A.BelloA.AyubaB. (2025). Effect of large language models artificial intelligence ChatGPT chatbot on achievement of computer education students. Educ. Inf. Technol.30 (9), 11863–11888. 10.1007/s10639-024-13293-8

  • CrossRef
  • Google Scholar

• 145

  MohebbiA. (2025). Enabling learner independence and self-regulation in language education using AI tools: a systematic review. Cogent Educ.12 (1), 2433814. 10.1080/2331186X.2024.2433814

  • CrossRef
  • Google Scholar

• 146

  MostafaR. B.KasamaniT. (2022). Antecedents and consequences of chatbot initial trust. Eur. J. Mark.56 (6), 1748–1771. 10.1108/EJM-02-2020-0084

  • CrossRef
  • Google Scholar

• 147

  MouakketS. (2020). Investigating the role of mobile payment quality characteristics in the United Arab Emirates: implications for emerging economies. Int. J. Bank Mark.38 (7), 1465–1490. 10.1108/IJBM-03-2020-0139

  • CrossRef
  • Google Scholar

• 148

  MoussawiS.KoufarisM. (2019). “Perceived intelligence and perceived anthropomorphism of personal intelligent agents: scale development and validation,” in Paper Presented at the Hawaii International Conference on System Sciences (HICSS). Maui: HI, Available online at:http://hdl.handle.net/10125/59452

  • Google Scholar

• 149

  MulyonoJ. A.SfenriantoS. (2022). Evaluation of customer satisfaction on Indonesian banking chatbot services during the COVID-19 pandemic. CommIT J.16 (1), 69–85. 10.21512/commit.v16i1.7813

  • CrossRef
  • Google Scholar

• 150

  MurtarelliG.CollinaC.RomentiS. (2023). ‘Hi! How can I help you today?’: Investigating the quality of chatbots–millennials relationship within the fashion industry. TQM J.35 (3), 719–733. 10.1108/TQM-01-2022-0010

  • CrossRef
  • Google Scholar

• 151

  NagashreeB. R. S.KrishnanR. V.RachayyanavarS. S.KumarA.KrishnaA. V. H. (2025). “Customer-centric technology: examining chatbot performance and fostering trust in the digital,” in Harnessing AI, Machine Learning, and IoT for Intelligent Business: Volume 1, eds. HamdanAllamBraendleUdo (Cham: Springer Nature Switzerland), 867–886. 10.1007/978-3-031-67890-5_78

  • CrossRef
  • Google Scholar

• 152

  NascimentoB.OliveiraT.TamC. (2018). Wearable technology: what explains continuance intention in smartwatches?J. Retail. Consum. Serv.43, 157–169. 10.1016/j.jretconser.2018.03.017

  • CrossRef
  • Google Scholar

• 153

  NgoT. T. A.PhanT. Y. N.NguyenT. K.LeN. B. T.NguyenN. T. A.LeT. T. D. (2025). Understanding continuance intention toward the use of AI chatbots in customer service among generation Z in Vietnam. Acta Psychol.259, 105468. 10.1016/j.actpsy.2025.105468

  • CrossRef
  • Google Scholar

• 154

  NguyenD. M.ChiuY.-T. H.LeH. D. (2021). Determinants of continuance intention towards Banks’ chatbot services in Vietnam: a necessity for sustainable development. Sustainability.13 (14), 7625. 10.3390/su13147625

  • CrossRef
  • Google Scholar

• 155

  NguyenT. H.LeX. C. (2025). Artificial intelligence-based chatbots–a motivation underlying sustainable development in banking: standpoint of customer experience and behavioral outcomes. Cogent Bus. Manag.12 (1), 2443570. 10.1080/23311975.2024.2443570

  • CrossRef
  • Google Scholar

• 156

  NicolaouA. I.MasonerM. M. (2013). Sample size requirements in structural equation models under standard conditions. Int. J. Account. Inf. Syst.14 (4), 256–274. 10.1016/j.accinf.2013.11.001

  • CrossRef
  • Google Scholar

• 157

  NielsenB. B.RaswantA. (2018). The selection, use, and reporting of control variables in international business research: a review and recommendations. J. World Bus.53 (6), 958–968. 10.1016/j.jwb.2018.05.003

  • CrossRef
  • Google Scholar

• 158

  NilashiM.IbrahimO.MirabiV. R.EbrahimiL.ZareM. (2015). The role of security, design and content factors on customer trust in Mobile commerce. J. Retail. Consum. Serv.26, 57–69. 10.1016/j.jretconser.2015.05.002

  • CrossRef
  • Google Scholar

• 159

  OkonkwoC. W.Ade-IbijolaA. (2021). Chatbots applications in education: a systematic review. Comput. Educ.: Artif. Intell.2, 100033. 10.1016/j.caeai.2021.100033

  • CrossRef
  • Google Scholar

• 160

  PalD.RoyP.ArpnikanondtC.ThapliyalH. (2022). The effect of trust and its antecedents towards determining users’ behavioral intention with voice-based consumer electronic devices. Heliyon.8 (4), e09271. 10.1016/j.heliyon.2022.e09271

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 161

  ParaskeviG.SaprikisV.AvlogiarisG. (2023). Modeling nonusers’ behavioral intention towards mobile chatbot adoption: an extension of the UTAUT2 model with mobile service quality determinants. Hum. Behav. Emerg. Technol.2023 (1), 8859989. 10.1155/2023/8859989

  • CrossRef
  • Google Scholar

• 162

  PearlC. (2016). Designing Voice User Interfaces: Principles of Conversational Experiences. O’Reilly Media, Inc.

  • Google Scholar

• 163

  PhamT. C.NgoT. T. A. (2026). AI tutor-based language learning: linking service quality to learners’ continuance intention through a dual-pathway model. J. Inf. Technol. Educ.: Res.25, 7. 10.28945/572

  • CrossRef
  • Google Scholar

• 164

  PillaiR.SivathanuB. (2020). Adoption of AI-based chatbots for hospitality and tourism. Int. J. Contemp. Hosp. Manag.32 (10), 3199–3226. 10.1108/IJCHM-04-2020-0259

  • CrossRef
  • Google Scholar

• 165

  PolyportisA. (2024). A longitudinal study on artificial intelligence adoption: understanding the drivers of ChatGPT usage behavior change in higher education. Front. Artif. Intell.6, 1324398. 10.3389/frai.2023.1324398

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 166

  PrasetyaS. A.ErwinA.GaliniumM. (2018). “Implementing Indonesian language chatbot for ecommerce site using artificial intelligence markup language (AIML),” in Proceedings of the Prosiding Seminar Nasional Pakar, Jakarta, Indonesia, 1 March 2018, 313–322.

  • Google Scholar

• 167

  RadziwillN. M.BentonM. C. (2017). Evaluating quality of chatbots and intelligent conversational agents. arXiv [Preprint]. Available online at:https://arxiv.org/abs/1704.04579

  • Google Scholar

• 168

  RafiqF.DograN.AdilM.WuJ.-Z. (2022). Examining consumer’s intention to adopt AI-chatbots in tourism using partial least squares structural equation modeling method. Math.10 (13), 2190. 10.3390/math10132190

  • CrossRef
  • Google Scholar

• 169

  RahimN. I. M.IahadN. A.YusofA. F.Al-SharafiM. A. (2022). AI-based chatbots adoption model for higher-education institutions: a hybrid PLS-SEM-neural network modelling approach. Sustainability.14 (19), 12726. 10.3390/su141912726

  • CrossRef
  • Google Scholar

• 170

  RahmanA. M.Al MamunA.IslamA. (2017). “Programming challenges of chatbot: current and future prospective,” in 2017 IEEE Region 10 Humanitarian Technology Conference (R10-HTC), (IEEE), 75–78.

  • Google Scholar

• 171

  RajaobelinaL.TepS. P.ArcandM.RicardL. (2021). Creepiness: its antecedents and impact on loyalty when interacting with a chatbot. Psychol. Mark.38 (12), 2339–2356. 10.1002/mar.21548

  • CrossRef
  • Google Scholar

• 172

  ReseA.GansterL.BaierD. (2020). Chatbots in retailers’ customer communication: how to measure their acceptance?J. Retail. Consum. Serv.56, 102176. 10.1016/j.jretconser.2020.102176

  • CrossRef
  • Google Scholar

• 173

  RocaJ. C.ChiuC.-M.MartínezF. J. (2006). Understanding e- learning continuance intention: an extension of the technology acceptance model. Int. J. Hum. Comput. Stud.64 (8), 683–696. 10.1016/j.ijhcs.2006.01.003

  • CrossRef
  • Google Scholar

• 174

  RoyR.AkhtarF.DasN. (2017). Entrepreneurial intention among science & technology students in India: extending the theory of planned behavior. Int. Entrep. Manag. J.13 (4), 1013–1041. 10.1007/s11365-017-0434-y

  • CrossRef
  • Google Scholar

• 175

  RuanY.MezeiJ. (2022). When do AI chatbots lead to higher customer satisfaction than human frontline employees in online shopping assistance? Considering product attribute type. J. Retail. Consum. Serv.68, 103059. 10.1016/j.jretconser.2022.103059

  • CrossRef
  • Google Scholar

• 176

  SahariY.Talib Al-KadiA. M.AliJ. K. M. (2023). A cross sectional study of ChatGPT in translation: magnitude of use, attitudes, and uncertainties. J. Psycholinguist. Res.52 (6), 2937–2954. 10.1007/s10936-023-10031-y

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 177

  Saldivar EspejoN. G.KassayS. M. V.CuellarH. M. (2025). “Determinants of satisfaction and initial trust in chatbots and their impact on customer engagement in the private healthcare sector,” in Marketing and Smart Technologies, eds. ReisJosé LuísGomesLuís MendesBogdanovićZoricaMarques dos SantosJosé Paulo (Singapore: Springer Nature), 211–222. 10.1007/978-981-96-3077-6_12

  • CrossRef
  • Google Scholar

• 178

  SandsS.FerraroC.CampbellC.TsaoH.-Y. (2021). Managing the human–chatbot divide: how service scripts influence service experience. J. Serv. Manag.32 (2), 246–264. 10.1108/JOSM-06-2019-0203

  • CrossRef
  • Google Scholar

• 179

  SelamatM. A.WindasariN. A. (2021). Chatbot for SMEs: integrating customer and business owner perspectives. Technol. Soc.66 (6), 101685. 10.1016/j.techsoc.2021.101685

  • CrossRef
  • Google Scholar

• 180

  SeoK. H.LeeJ. H. (2021). The emergence of service robots at restaurants: integrating trust, perceived risk, and satisfaction. Sustainability.13 (8), 4431. 10.3390/su13084431

  • CrossRef
  • Google Scholar

• 181

  ShahzadM. F.XuS.AnX.JavedI. (2024). Assessing the impact of AI-chatbot service quality on user e-brand loyalty through chatbot user trust, experience and electronic word of mouth. J. Retail. Consum. Serv.79, 103867. 10.1016/j.jretconser.2024.103867

  • CrossRef
  • Google Scholar

• 182

  ShannonC. E. (1955). A Proposal for the Dartmouth Summer Research Project on Artificial Intelligence. Hanover, New Hampshire: John McCarthy, Department of Mathematics, Dartmouth College.

  • Google Scholar

• 183

  ShawarB. A. (2017). Integrating CALL systems with chatbots as conversational partners. Comput. Sist.21 (4), 615–626.

  • Google Scholar

• 184

  SheehanB.JinH. S.GottliebU. (2020). Customer service chatbots: anthropomorphism and adoption. J. Bus. Res.115, 14–24. 10.1016/j.jbusres.2020.04.030

  • CrossRef
  • Google Scholar

• 185

  ShevlinM.MilesJ. N. (1998). Effects of sample size, model specification and factor loadings on the GFI in confirmatory factor analysis. Personal. Individ. Differ. 25, 85–90.

  • Google Scholar

• 186

  ShiY.ZhanW.LinC. (2025). The impact of hotel robots’ service quality on continuance intention: the moderating effect of personal innovation. Front. Robot. AI.12, 1667123. 10.3389/frobt.2025.1667123

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 187

  ShinD. (2021). The effects of explainability and causability on perception, trust, and acceptance: implications for explainable AI. Int. J. Hum. Comput. Stud.146, 102551. 10.1016/j.ijhcs.2020.102551

  • CrossRef
  • Google Scholar

• 188

  SilvaF. A.ShojaeiA. S.BarbosaB. (2023). Chatbot-based services: a study on customers’ reuse intention. J. Theor. Appl. Electron. Commer. Res.18 (1), 457–474. 10.3390/jtaer18010024

  • CrossRef
  • Google Scholar

• 189

  SilvaG. R. S.CanedoE. D. (2024). Towards user-centric guidelines for chatbot conversational design. Int. J. Hum.–Comput. Interact.40 (2), 98–120. 10.1080/10447318.2022.2118244

  • CrossRef
  • Google Scholar

• 190

  SitthiponT.SiripipatthanakulS.PhayaphromB.SiripipattanakulS.LimnaP. (2022). Determinants of customers’ intention to use healthcare chatbots and apps in Bangkok, Thailand. Int. J. Behav. Anal.2 (2), 1–15.

  • Google Scholar

• 191

  SkantzeG. (2007). Error Handling in Spoken Dialogue Systems-Managing Uncertainty, Grounding and Miscommunication. (Ph.D. thesis). KTH Stockholm, Stockholm, Sweden. (Computer Science and Communication; Department of Speech,Music and Hearing).

  • Google Scholar

• 192

  SobaihA. E. E.ElshaerI. A.HasaneinA. M. (2024). Examining students’ acceptance and use of ChatGPT in Saudi Arabian higher education. Eur. J. Investig. Health Psychol. Educ.14 (3), 709–721. 10.3390/ejihpe14030047

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 193

  SohailS. S.FarhatF.HimeurY.NadeemM.MadsenD. Ø.SinghY.et al (2023). Decoding ChatGPT: a taxonomy of existing research, current challenges, and possible future directions. J. King Saud Univ. Comput. Inf. Sci.35 (8), 101675. 10.1016/j.jksuci.2023.101675

  • CrossRef
  • Google Scholar

• 194

  StöhrC.OuA. W.MalmströmH. (2024). Perceptions and usage of AI chatbots among students in higher education across genders, academic levels and fields of study. Comput. Educ.: Artif. Intell.7, 100259.

  • Google Scholar

• 195

  StrzeleckiA. (2024). Students’ acceptance of ChatGPT in higher education: an extended unified theory of acceptance and use of technology. Innov. High. Educ.49 (2), 223–245. 10.1007/s10755-023-09686-1

  • CrossRef
  • Google Scholar

• 196

  SundjajaA. M.UtomoP.CollineF. (2025). The determinant factors of continuance use of customer service chatbot in Indonesia e-commerce: extended expectation confirmation theory. J. Sci. Technol. Policy Manag.16 (1), 182–203. 10.1108/JSTPM-04-2024-0137

  • CrossRef
  • Google Scholar

• 197

  TeoT. S.SrivastavaS. C.JiangL. I. (2008). Trust and electronic government success: an empirical study. J. Manag. Inf. Syst.25 (3), 99–132. 10.2753/MIS0742-1222250303

  • CrossRef
  • Google Scholar

• 198

  TianY.WangX. (2022). A study on psychological determinants of users’ autonomous vehicles adoption from anthropomorphism and UTAUT perspectives. Front. Psychol.13, 986800. 10.3389/fpsyg.2022.986800

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 199

  TiwariC. K.BhatM.KhanS. T.SubramaniamR.KhanM. A. I. (2024). What drives students toward ChatGPT? An investigation of the factors influencing adoption and usage of ChatGPT. Interact. Technol. Smart Educ.21 (3), 333–355. 10.1108/ITSE-04-2023-0061

  • CrossRef
  • Google Scholar

• 200

  TrawnihA.Al-MasaeedS.AlsoudM.AlkufahyA. (2022). Understanding artificial intelligence experience: a customer perspective. Int. J. Data Netw. Sci.6 (4), 1471–1484. 10.5267/j.ijdns.2022.5.004

  • CrossRef
  • Google Scholar

• 201

  TrivediJ. (2019). Examining the customer experience of using banking chatbots and its impact on brand love: the moderating role of perceived risk. J. Internet Commer.18 (1), 91–111. 10.1080/15332861.2019.1567188

  • CrossRef
  • Google Scholar

• 202

  TrivediR.PattusamyM. (2022). Performance pressure and innovative work behaviour: the role of problem-orientated daydreams. IIMB Manag. Rev.34 (4), 333–345. 10.1016/j.iimb.2022.12.005

  • CrossRef
  • Google Scholar

• 203

  TuringI. B. A. (1950). Computing machinery and intelligence-AM turing. Mind.59 (236), 433–460. 10.1093/mind/LIX.236.433

  • CrossRef
  • Google Scholar

• 204

  TurpeningE.LittletonA. (2016). “The 2016 State of Social Business” Social’s Shift From Innovator to Integrator. Altimeter Group.

  • Google Scholar

• 205

  UrbachN.MüllerB. (2011). “The updated DeLone and McLean model of information systems success,” in Information Systems Theory: Explaining and Predicting Our Digital Society, Vol. 1, (New York, NY: Springer), 1–18.

  • Google Scholar

• 206

  UsakliA.RasoolimaneshS. M. (2023). “Which SEM to use and what to report? A comparison of CB-SEM and PLS-SEM,” in Cutting Edge Research Methods in Hospitality and Tourism, eds. OkumusFevziRasoolimaneshS. MostafaJahaniShiva, (Emerald publishing Limited), 5–28. 10.1108/978-1-80455-063-220231002

  • CrossRef
  • Google Scholar

• 207

  WheatonB.MuthenB.AlwinD. F.SummersG. F. (1977). Assessing reliability and stability in panel models. Sociol. Methodol. 8, 84–136

  • Google Scholar

• 208

  WuenderlichN. V.PaluchS. (2017). “A nice and friendly chat with a bot: user perceptions of AI-based service agents,” in ICIS 2017 Proceedings, 11. Available online at:https://aisel.aisnet.org/icis2017/ServiceScience/Presentations/11

  • Google Scholar

• 209

  WutT. M.LeeS. W.XuJ.KwokM. L. J. (2025). Do trusting belief and social presence matter? Service satisfaction in using AI chatbots: necessary condition analysis and importance-performance map analysis. Informatics.12 (3), 91. 10.3390/informatics12030091

  • CrossRef
  • Google Scholar

• 210

  YangM.PengX.WangQ.ZhaoY. C.WangX. (2025). Is being human-like beneficial? The effect of anthropomorphism on chatbot persuasion in e-commerce. Internet Res., (ahead of print). 10.1108/INTR-10-2023-0866

  • CrossRef
  • Google Scholar

• 211

  YangS.LeeY. J. (2017). The dimensions of M-interactivity and their impacts in the mobile commerce context. Int. J. Electron. Commer.21 (4), 548–571. 10.1080/10864415.2016.1355645

  • CrossRef
  • Google Scholar

• 212

  Yildiz DurakH.OnanA. (2024). Predicting the use of chatbot systems in education: a comparative approach using PLS-SEM and machine learning algorithms. Curr. Psychol.43 (28), 23656–23674. 10.1007/s12144-024-06072-8

  • CrossRef
  • Google Scholar

• 213

  YoonH. S.OccenaL. G. (2015). Influencing factors of trust in consumer-to- consumer electronic commerce with gender and age. Int. J. Inf. Manage.35 (3), 352–363. 10.1016/j.ijinfomgt.2015.02.003

  • CrossRef
  • Google Scholar

• 214

  YuC. C.YanJ. Z.CaiN. (2024). ChatGPT in higher education: factors influencing ChatGPT user satisfaction and continued use intention. Front. Educ.9, 1354929. 10.3389/feduc.2024.135492

  • CrossRef
  • Google Scholar

• 215

  YuH. (2023). Reflection on whether chat GPT should be banned by academia from the perspective of education and teaching. Front. Psychol.14, 1181712. 10.3389/fpsyg.2023.1181712

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 216

  ZafarA. U.QiuJ.LiY.WangJ.ShahzadM. (2021). The impact of social media celebrities’ posts and contextual interactions on impulse buying in social commerce. Comput. Human. Behav.115, 106178. 10.1016/j.chb.2019.106178

  • CrossRef
  • Google Scholar

• 217

  ZhangR. W.LiangX.WuS.-H. (2024a). When chatbots fail: exploring user coping following a chatbots-induced service failure. Inf. Technol. People.37 (8), 175–195. 10.1108/ITP-08-2023-0745

  • CrossRef
  • Google Scholar

• 218

  ZhangR.ZouD.ChengG. (2024b). A review of chatbot-assisted learning: pedagogical approaches, implementations, factors leading to effectiveness, theories, and future directions. Interact. Learn. Environ.32 (8), 4529–4557. 10.1080/10494820.2023.2202704

  • CrossRef
  • Google Scholar

• 219

  ZhangS.HuZ.LiX.RenA. (2022). The impact of service principal (service robot vs. human staff) on service quality: the mediating role of service principal attribute. J. Hosp. Tour. Manag.52, 170–183. 10.1016/j.jhtm.2022.06.014

  • CrossRef
  • Google Scholar

• 220

  ZhangW.ZhangW.WangC.DaimT. U. (2022). What drives continuance intention of disruptive technological innovation? The case of e-business microcredit in China. Technol. Anal. Strateg. Manag.34 (8), 905–918. 10.1080/09537325.2021.1932798

  • CrossRef
  • Google Scholar