Diana Yanet Gaviria Rodríguez1†
Alejandro Valencia-Arias2*†
Juan Guillermo Arango Arango3†
Juan Manuel Raunelli Sander4†
Lilian Verde Flores2†
Jackeline Valencia5†- 1Facultad de Ciencias Económicas y Administrativas, Instituto Tecnológico Metropolitano, Medellín, Colombia
- 2Escuela de Ingeniería Industrial, Universidad Señor de Sipán, Chiclayo, Peru
- 3Ciencias Básicas, Instituto Tecnológico Metropolitano, Medellín, Colombia
- 4Facultad de Ingeniería Arquitectura y Urbanismo, Universidad Señor de Sipán, Chiclayo, Peru
- 5Vicerrectoría de Investigación y postgrado, Universidad de Los Lagos, Osorno, Chile
Introduction: This study examines the integration of Artificial Intelligence (AI)-based tools into university-level accounting education in Medellín, Colombia. The objective was to identify the factors that influence students' intention to adopt these technologies, using theoretical frameworks such as the Theory of Planned Behavior (TPB) and the Technology Acceptance Model (TAM).
Methods: An experimental methodology was applied, which included the development of an educational video incorporating AI and an accounting simulator. A total of 105 students participated in the study and completed a Likert-type questionnaire designed to evaluate constructs including attitude, perceived usefulness, ease of use, subjective norm, and behavioral control.
Results: The analysis revealed that perceived ease of use significantly influenced both perceived usefulness and behavioral control. Additionally, subjective norm had an impact on attitude and intention to use. However, perceived usefulness did not translate into favorable attitudes toward AI adoption, indicating a gap between the functionalities students recognized and their expectations of the technology.
Discussion: The findings highlight the importance of contextualizing AI functionalities within educational settings. They suggest the need for pedagogical strategies that align technological tools with students' expectations and foster a more receptive environment toward digital innovation in accounting education.
1 Introduction
Technological advances have profoundly transformed educational processes, enabling the integration of innovative tools into university curricula. Among these emerging technologies, Artificial Intelligence (AI) has proven to be one of the most promising, enhancing efficiency, personalizing learning experiences, and preparing students to meet the demands of the contemporary labor market (Gruenbichler et al., 2024). The advent of the digital revolution has prompted higher education institutions to adopt AI-based solutions, recognizing their potential to automate processes, deliver real-time analytics, and enrich the learning experience.
In the context of higher education, Artificial Intelligence (AI) has begun to play a pivotal role. This is evidenced by its enabling of personalized learning experiences, identification of knowledge gaps, and enhancement of interaction through virtual tutors and intelligent assessment systems (Norzelan et al., 2024). Its integration into academic programs addresses two crucial needs: the need for a more technologically proficient workforce and the development of digital and adaptive skills that are critical in today's era. This transformation has been particularly evident in various academic disciplines, including public accounting.
In this context, AI has proven to be a valuable tool in the training of future accountants, supporting the management of complex tasks such as analyzing large volumes of data, forecasting financial trends, and optimizing accounting processes (Al-Okaily et al., 2023). The integration of these technologies has been instrumental in enhancing the quality of accounting education and preparing students to thrive in a business environment that is increasingly characterized by automation and digital transformation.
However, the effective adoption of AI in university education is influenced by various factors, including the availability of technological infrastructure, the level of digital literacy among students and faculty, perceptions of AI's usefulness, and institutional policies that support its implementation (Algerafi et al., 2023). In the specific case of accounting students, it is essential to understand the factors that influence their intention to adopt such technologies. In contexts where traditional methods have historically prevailed, variables such as perceived usefulness, ease of use, trust in the technology, and willingness to learn new digital tools are particularly relevant (Bui et al., 2025).
In light of these considerations, this study aims to identify the factors that influence the intention to use artificial intelligence in the educational activities of accounting students in Medellín, Colombia. The objective is to contribute knowledge that can enhance the implementation of these technologies in academic programmes, promoting accounting education that is more aligned with the demands of the twenty-first century.
1.1 Theoretical framework
A number of studies have analyzed the adoption of AI in higher education, identifying key factors that influence its implementation. Specifically, Alhumaid et al. (2023) emphasize that perceived compatibility and ease of use are pivotal factors in determining the adoption of AI applications in educational institutions. In contrast, Pillai et al. (2024) emphasize the significance of variables such as personalization, interactivity, and trust in the utilization of AI-based teaching bots (T-bots) within higher education. In contrast, authors such as Alotaibi and Alshehri (2023) have emphasized that the integration of AI in higher education in Saudi Arabia presents opportunities for the transformation of teaching methodologies. However, challenges related to teacher training and technological infrastructure have also been identified.
In the specific context of accounting and finance programs, research conducted by Sudaryanto et al. (2023) examined the factors influencing the adoption of AI technology by accounting students. The study revealed that ease of use and perceived usefulness are significant factors in AI adoption, while digital competence and technological readiness did not demonstrate a direct impact. In a similar vein, Al-Okaily (2024) conducted a study that analyzed the impact of AI on data quality and financial reporting in the field of accounting and financial disclosure. The study underscored the significance of adopting digital disclosure practices in enhancing data quality.
In order to comprehend adoption factors from a behavioral perspective, one must first acquaint oneself with the theoretical underpinnings of the Planned Behavior Model (TPB). This seminal model, proposed by Ajzen (1991), posits that an individual's inclination to engage in a particular behavior is influenced by three pivotal factors: attitude toward the behavior, subjective norms, and perceived control over the behavior. In this context, attitude is defined as a person's positive or negative evaluation of an action. Subjective norms are defined as the influence of others on an individual's decision, while perceived control addresses the ease or difficulty an individual perceives in performing the behavior.
In contrast, the Technology Acceptance Model (TAM), proposed by Davis et al. (1989), expands this perspective by focusing on two key factors: perceived usefulness and perceived ease of use. The concept of perceived usefulness is delineated as the perception that a technology will enhance task performance, while perceived ease of use is defined as the perception that utilizing the technology will require minimal effort. These factors are considered fundamental, as they directly influence users' attitudes toward technology. The adoption of such technology is contingent upon its perceived usefulness and ease of use by the users, which in turn fosters a positive attitude toward toward its utilization. Consequently, this has been demonstrated to enhance their inclination to adopt it, underscoring the significance of these two factors in technology acceptance models.
Integrating the TPB and TAM facilitates a more comprehensive understanding of technology adoption by combining the factors of perceived usefulness and ease of use (TAM) with social norms and perceived control over behavior (TPB). Gaviria et al. (2022) employed this integration to examine university students' intention to utilize virtual learning objects. Their findings demonstrated that attitudes toward technology, social influences, and perceived control are pivotal determinants of technology adoption.
1.2 Perceived usefulness
Perceived usefulness is a pivotal factor in technology acceptance models, as it pertains to the extent to which a user perceives that utilizing a technology will enhance their performance in a particular task. This variable has been demonstrated to exert a direct influence on attitudes toward technology, with a concomitant positive impact on users' propensity to adopt it. As posited by Margraf et al. (2020), perceived usefulness is foundational for users to cultivate a favorable attitude toward technology. This phenomenon can be attributed to the principle that users are more likely to be motivated to utilize a tool if they believe it will be beneficial to their daily activities. Recent studies have corroborated this association: Wiprayoga et al. (2023) found that perceived usefulness significantly enhances attitude, which in turn mediates the adoption process. Denovan and Marsasi (2025) demonstrated that for Generation Y and Z consumers, perceived usefulness not only affects attitude directly but also indirectly through satisfaction and trust.
With regard to behavioral intention, perceived usefulness is also of crucial importance, in that users tend to exhibit a greater intention to adopt a technology if they believe its use will generate clear benefits. The relationship between perceived usefulness and behavioral intention has been extensively validated in prior studies, including that of Humida et al. (2022), who determined that perceived usefulness is a primary predictor of intention to use. Amoako-Gyampah's (2007) research yielded analogous results, emphasizing that perceived usefulness, particularly when coupled with user engagement, directly reinforces in ERP adoption. In recent educational contexts, Alshammari and Babu (2025) demonstrated that perceived usefulness positively influences students' intention to use AI-powered tools like ChatGPT, with satisfaction acting as a mediating factor. Consequently, it can be posited that perceived usefulness exerts a positive influence on both attitude toward behavior and behavioral intention, although its impact may be modulated by contextual and psychological variables such as satisfaction, trust, and user expectations.
H1: Perceived usefulness has a positive influence on attitude toward behavior.
H2: Perceived usefulness has a positive influence on behavioral intention.
1.3 Perceived ease of use
The notion of perceived ease of use is a widely acknowledged critical factor in the adoption of novel technologies. This is because it refers to the extent to which users believe that using a given system will require minimal effort. According to Tahar et al. (2020), when users perceive a technology as easy to use, this perception tends to enhance their evaluation of its usefulness, as they associate ease with a more efficient and less demanding experience. Alshammari and Babu's (2025) study demonstrated that the ease of use of Chat GPT significantly influences students‘ behavioral intention to adopt it, with satisfaction acting as a mediating factor. Kim et al. (2025) similarly found that ease of use enhances both perceived usefulness and adoption, especially when paired with enjoyment. Cao et al. (2025) demonstrated that in rural contexts, ease of use becomes paramount when coupled with adequate training and digital literacy. As posited by Oematan et al. (2024), user satisfaction exerts a mediating role between ease of use and behavioral intention in e-commerce settings. The findings, when considered collectively, suggest that ease of use not only enhances users' perception of utility but also plays a pivotal role in the emotional and motivational dimensions that facilitate adoption.
Furthermore, perceived ease of use has been demonstrated to exert a positive influence on perceived behavioral control. When users interact with technologies that are intuitive and straightforward, they report greater confidence in their ability to use them effectively. Valencia-Arias et al. (2023) posit that this perceived autonomy is fundamental to promoting self-efficacy in technology adoption. Hoque et al. (2024) further corroborate this finding, demonstrating that perceived behavioral control serves as a mediator between innovativeness and the continued usage of e-money platforms. In the context of AI-assisted language learning, Wu et al. (2024) found that ease of use directly enhanced students' sense of control and willingness to engage with digital tools. Finally, Warsono et al. (2023) assert that perceived ease of use exerts a consistent and positive influence on perceived usefulness, perceived behavioral control, and behavioral intention. A synthesis of extant studies indicates that perceived ease of use not only serves as a predictor of user intention but also constitutes a foundational component of perceived capability in digital environments.
H3: Perceived ease of use has a positive influence on perceived usefulness.
H4: Perceived ease of use has a positive influence on perceived behavioral control.
H5: Perceived ease of use has a positive influence on behavioral intention.
1.4 Attitude toward behavior
Attitude toward behavior is defined as an individual's positive or negative evaluation of performing a specific action, such as the adoption of a new technology. Research has demonstrated that an individual's propensity to engage with a given technology is contingent upon their disposition toward it. In essence, a positive attitude toward a technology is associated with increased inclination to engage with it. This heightened level of engagement has been shown to foster an increased likelihood of adoption. This assertion is corroborated by the findings of Ho et al. (2020), who determined that positive attitudes toward a given behavior are a significant predictor of the intention to engage in that behavior. In a similar vein, Bai et al. (2024) discovered that teachers' attitudes toward technology significantly mediate the relationship between their technology self-efficacy and behavioral intention, underscoring the pivotal role of attitude in determining actual adoption. Furthermore, Zhao et al. (2025) demonstrated in the context of tourism behavior that attitude influences intention directly and also moderates the strength of other psychological factors, such as motivation and perceived value. Consequently, if users find a technology beneficial and feel at ease with its use, their propensity to adopt it is notably heightened, thereby streamlining the decision-making process toward its effective implementation.
H6: Attitude toward behavior has a positive influence on behavioral intention.
1.5 Subjective norm
The subjective norm, defined as an individual's perception of social expectations and the influence of others on their decisions, plays a crucial role in shaping attitudes and behaviors. As La Barbera and Ajzen (2020) demonstrate, when individuals perceive that their acquaintances hold a belief in the adoption of a given technology, they are more inclined to adopt a favorable attitude toward its utilization. This social pressure functions as a motivating factor, as individuals often align themselves with the expectations of their social environment, thereby reinforcing their propensity to adopt new technologies. Baba et al. (2025) corroborate the notion that subjective norms serve as a substantial predictor of behavioral intention, a phenomenon that is particularly pronounced when these norms are firmly entrenched within a robust community ethos.
In a similar vein, Baba et al. (2025) study found that perceived social expectations significantly influence individuals‘ attitudes and willingness to engage in health-related behaviors, such as vaccination. This highlights the broader role of subjective norms in behavior regulation. In the context of AI adoption, Wang et al. (2024) demonstrated that social influence, as operationalized through subjective norm, significantly affects university students' intentions to use generative artificial intelligence tools.
Furthermore, subjective norms have been demonstrated to influence an individual's perception of their ability to control their behavior, particularly in the context of technology use. As posited by Aschwanden et al. (2021), individuals experience an increase in confidence in their ability to adopt a specific technology when they perceive societal approval of its utilization. This phenomenon is further exacerbated by the perceived support derived from one's immediate social circle. Essam et al. (2025) discovered that subjective norms, in conjunction with perceived behavioral control, serve a dual moderating function in shaping behavioral intentions, particularly in consumption contexts.
Similar vein, Kaba et al. (2025) identified that social influence interacts with demographic factors to shape knowledge-sharing intentions through both attitudinal and normative mechanisms. This social effect engenders a heightened sense of control, prompting individuals to align their actions with the expectations of their peers. Wismantoro and Susilowati (2025) also emphasize that subjective norms, when reinforced by positive attitudes and control beliefs, enhance pro-environmental behavioral intentions. Finally, given the potential of subjective norm to influence both attitude and perceived behavioral control, it is conceivable that it could exert a direct effect on the intention to adopt the technology, particularly in contexts where individuals not only perceive themselves as capable and hold favorable attitudes, but also feel socially encouraged to act accordingly (La Barbera and Ajzen, 2020).
H7: The subjective norm has a positive influence on attitude toward behavior.
H8: The subjective norm has a positive influence on perceived behavioral control.
H9: The subjective norm has a positive influence on behavioral intention.
1.6 Control of perceived behavior
The perception of control over behavior plays a fundamental role in decision-making, as it directly influences a person's ability to act in accordance with their intentions. The propensity for individuals to adopt new technologies is positively correlated with their perceived access to the requisite resources, skills, and support systems. This sense of self-efficacy reinforces their confidence that they can overcome any obstacles that arise, which in turn motivates them to take the desired action (Vamvaka et al., 2020). Consequently, an augmented perception of autonomy in behavior engenders a heightened probability of individuals actualizing their intentions to adopt the technology or behavior in question.
H10: Perceived behavioral control has a positive influence on behavioral intention.
As illustrated in Figure 1, the proposed theoretical model integrates 10 hypotheses derived from the integration of TPB and TAM theories. The model offers a clear illustration of the relationships between key variables, including perceived usefulness, perceived ease of use, subjective norm, attitude toward the behavior, perceived behavioral control, and behavioral intention. It provides a coherent representation of the mutual influences that shape the technology adoption process.
Figure 1. Proposed model. Prepared by the authors.
2 Methodology
An experimental case study was conducted, applying a pedagogical strategy to a classroom training activity. The objective of this study was to ascertain the factors that influence the intention to utilize the “AI VIDEO COMBINED WITH THE SIMULATOR.” The present study concentrated on the evaluation of the relationship between constructs such as attitude toward the behavior, behavioral intention, perceived behavioral control, perceived ease of use, perceived usefulness, and subjective norms.
The present study was conducted in two parts. The initial phase entailed the development of an AI-generated video to illustrate the unit-based depreciation method. The script generated by Chat GPT was subjected to a process of review and modification, and was subsequently recorded using the “Fliki.ai” program. The simulator was then designed using Geo Gebra software, incorporating all the concepts of depreciation and impairment. These concepts and definitions were applied in accordance with the regulatory framework of Decree 2420 of 2015. It is evident that Annexes 2 and 3 of the International Financial Reporting Standards illustrate the recognition, initial and subsequent measurement of property, plant and equipment, and the standards that develop impairment. Subsequent to this, the relevant equations and algorithms were designed. Following the construction of the simulator, it was uploaded to the website in the form of a Descartes book (Gaviria et al., 2022). This website contains an open repository of works oriented to Interactive Learning Objects (ILOs). These simulators are designed with free software such as Geo Gebra and Descartes, along with video materials to guide learning for university students in various subjects, including Basic Mathematics, Differential Calculus, and Accounting Sciences.
In the second part of the study, a training activity was implemented in the classroom during the 2024–01 semester. The activity was entitled “VIDEO AI COMBINED WITH THE SIMULATOR”. The study was administered to five groups of students who had previously studied the topic “Depreciation Methods.” Of the five groups, three were taught by professors who were not involved in the study's research, while two groups were taught by one of the study's researchers. All participating students were enrolled at the Metropolitan Technological Institute (ITM) in Medellín, Colombia, with the majority belonging to the Finance Department and representing the Public Accounting and Technology in Cost and Budget Analysis programmes.
Prior to the implementation of the activity in the classroom, a preparatory and verification process was conducted with three professors other than the study's researchers. At that juncture, the activity was disseminated to the teachers, in the form of the simulator with an established guide. This enabled the teachers to familiarize themselves and conduct a verification test using their own exercises and methodology. This preliminary stage enabled the validation of the simulator and the methods used to guide the process step by step, ensuring that the fieldwork was conducted with reliable data and was carried out efficiently and in a coordinated manner to obtain valid and accurate research results.
Following the implementation of the training activity, an instrument adapted to this study was developed and validated in Factors That Affect the Usage Intention of Virtual Learning Objects by College Students (Gaviria et al., 2022). The questionnaire comprised 20 questions, designed to assess constructs including attitude toward the behavior, behavioral intention, perceived behavioral control, perceived ease of use, perceived usefulness, and subjective norms.
As part of the methodological design, the educational materials were developed by researchers D.Y.G-R. and J.G.A.A., who created an artificial intelligence-based video for pedagogical purposes. The content was initially generated using the Chat GPT tool and subsequently reviewed, adapted, and validated by three professors with experience in accounting and technologies applied to education. The validation by expert judgment was focused on ensuring conceptual clarity, alignment with the learning objectives, and the pedagogical relevance of the resource. The final video was produced using the Fliki.ai platform, ensuring a comprehensible and engaging visual format for students.
Furthermore, the interactive simulator was developed by J.G.A.A. and L.V.F. in the Geo Gebra environment, incorporating the regulatory principles established in Decree 2420 of 2015 on depreciation and impairment of assets. The simulator was integrated into an interactive digital book and was pilot-tested by three teachers who were not members of the research team. This preliminary phase facilitated the validation of the simulator's usability, the clarity of its instructions, and its practical functionality within authentic classroom contexts. The feedback received was integrated to improve the tool before its formal implementation in the participating groups.
The questions were designed using a Likert-type scale, which allowed respondents to indicate their level of agreement or disagreement with various statements. The scale ranged from 5 (strongly agree) to 1 (strongly disagree), thereby capturing the intensity of participants' perceptions. This rating system is commonly used in behavioral and educational research to quantify attitudes and opinions. The specific structure of the scale is presented in Table 1.
Table 1. Measurement scale used for likert-type questions.
However, subsequent to the validation of the model, it was determined that a mere 13 of these questions met the criteria for relevance for the analysis. This reduction ensured that the selected questions were the most effective for addressing the six constructs.
The completion of this instrument was achieved by 105 students under the following conditions: the students had studied the topic of Property, Plant, and Equipment —The Depreciation by Use Method and the Deterioration of Property, Plant, and Equipment. The students participated in a training activity with the instructor, during which they were explained using “VIDEO AI COMBINED WITH THE SIMULATOR.”
The data collected during this study will be treated with strict confidentiality and will not be shared with any external entity. The data will be utilized exclusively for research purposes, with analysis based on responses to the 20 Likert-type questions, age, gender, and semester completed. Confidentiality protection was ensured by the non-disclosure of the personal information of the students who participated in the study (Publication Manual of the American Psychological Association).
To complement the experimental implementation, it is imperative to clarify that the activity was structured in a controlled classroom environment, with a duration of ~90 m per session. The experimental groups received the same pedagogical sequence. First, students viewed the AI-generated video, followed by hands-on interaction with the simulator under the guidance of the researcher. Prior to the initiation of the study, the teachers received orientation on the instructional strategy to ensure fidelity of implementation across groups. During the activity, students were encouraged to explore the simulator individually or in pairs and to reflect on how the video supported their understanding of the topic. The researchers observed the sessions without interfering, ensuring consistency in the execution while minimizing bias. This implementation design enabled the structured observation of students' engagement with AI-based educational tools within a natural academic context.
The questionnaire utilized in this study was adapted from the validated instrument developed by Gaviria et al. (2022), titled Factors That Affect the Usage Intention of Virtual Learning Objects by College Students. The original instrument was designed to assess factors related to technology acceptance in educational contexts and was grounded in the Technology Acceptance Model (TAM) and the Theory of Planned Behavior (TPB). For the purposes of this study, the original items were reviewed and modified by the research team to ensure conceptual alignment with the specific constructs under investigation: attitude toward behavior, behavioral intention, perceived behavioral control, perceived ease of use, perceived usefulness, and subjective norms. In adapting the instrument, additional input was obtained from three subject-matter experts in accounting and educational research, who contributed to the refinement of item wording, content validity, and contextual relevance.
The development process entailed the explicit mapping of each question to a theoretical construct. For instance, items pertaining to perceived usefulness included statements such as “I believe the AI-based simulator helps me better understand depreciation methods,” while perceived behavioral control was assessed with items like “I feel confident in my ability to use AI-based tools for accounting tasks.” The clarity and internal consistency of the questions were assessed in a pilot test conducted with 20 students from a different cohort. This pilot test led to the refinement of several items prior to final data collection. As illustrated in Table 1 The Likert-type structure was maintained from the original scale to preserve psychometric integrity and ensure comparability with prior research.
3 Statistical model results
The statistical analysis was conducted using Structural Equation Modeling (SEM), a multivariate analysis technique that, according to Zeng et al. (2021), is considered both routine and essential in the validation of theoretical models. Specifically, this study employed the Partial Least Squares Structural Equation Modeling (PLS-SEM) approach, which focuses on maximizing the explained variance of the dependent latent constructs rather than reproducing a theoretical covariance matrix. The analysis was performed using Smart PLS 4 software (Ringle et al., 2022).
The analytical process comprised two main stages: the measurement model and the structural model. In the measurement model, the assessment included convergent validity (via factor loadings and Average Variance Extracted AVE), internal consistency reliability (using Cronbach's alpha and composite reliability), and discriminant validity. The latter was evaluated through the cross-loadings criterion, ensuring that each indicator loads more highly on its assigned construct than on any other, as recommended by Hair et al. (2019). In the structural model, the proposed hypotheses were tested, and the explanatory power of the model was examined using coefficients of determination (R2) and path coefficients.
3.1 Analysis of the measurement model
The analysis initiates with the validation of the model's external loadings to ensure convergent validity. According to experts in the field, external loadings should exceed 0.708, thereby ensuring that the square of that number indicates that the construct score incorporates a minimum of 50% of the variable's variance (Hair et al., 2017). In this sense, the initial criterion is fulfilled, as demonstrated in Table 2 which shows that all indicators exhibit satisfactory external loadings. Additionally, it has been observed that the correlations of the indicators' external loadings with respect to their construct are greater than the correlations with other constructs (which they should not measure) (Rasoolimanesh, 2022). Moreover, indicators that do not meet the specified criterion are eliminated in order to avoid any detrimental impact on the model's quality. Consequently, the following constructs are eliminated: The following elements must be given full consideration: The following elements are listed herewith: BI1, PU1, PBC1, PBC2, SN1, SN3, and SN5.
Table 2. Cross-loads of measurement items for each construct.
The aforementioned validation encompasses a multicollinearity analysis, which is conducted utilizing the variance inflation factor (VIF). In the context of PLS-SEM, a tolerance value of less than 0.2 is recommended, as values greater than 5 indicate critical levels of collinearity between constructs. Consequently, a value of less than 3.3 is proposed (Martínez Ávila and Fierro Moreno, 2018). The results of this analysis, presented in Table 3 demonstrate that this criterion is met and, therefore, the model does not present collinearity issues.
Table 3. Convergent validity of the model.
Furthermore, the convergent validity analysis demonstrates that the Average Variance Extracted (AVE) criterion is met, which is regarded as the most significant test. According to experts in the field, the AVE value for each construct must be a minimum of 0.5 in order to be considered acceptable (Purwanto and Sudargini, 2021). The results presented in Table 3 further corroborate the validity of this criterion. Consequently, the model demonstrates convergent validity.
A reliability test is also conducted, utilizing composite reliability (CR) to ascertain the model's internal consistency. The findings of this study indicate that the model demonstrates sufficient internal consistency. Consequently, higher CR values are indicative of greater reliability and internal consistency. As posited by Hair et al. (2019), reliability values ranging from 0.6 to 0.7 are deemed to be “acceptable in exploratory research,” whereas values between 0.70 and 0.90 are classified as “satisfactory to good.”
In this study, CR was prioritized over Cronbach's alpha, another reliability measure, as CR has been demonstrated to produce higher values than Cronbach's alpha (Purwanto and Sudargini, 2021). This discrepancy can be attributed to the fact that CR, in contrast to Cronbach's alpha, does not presuppose that all indicators are equally reliable. Consequently, CR is more appropriate for PLS-SEM (Hair et al., 2011). Consequently, the outcomes presented in Table 3 substantiate that the model meets the established reliability criteria.
Furthermore, the instrument's psychometric properties were evaluated using reliability and validity indicators. The reliability of the data was assessed using Cronbach's alpha coefficient, a statistical measure of internal consistency. The coefficient values for all constructs exceeded the 0.70 threshold, indicating acceptable reliability. The Composite Reliability (CR) metric was also utilized, yielding values ranging from 0.78 to 0.86, which exceed the recommended minimum of 0.70. The external factor loadings of the items were found to be above 0.708, a criterion that ensures the convergent validity of the indicators. This validity is further substantiated by the Average Variance Extracted (AVE) values, which consistently exceeded 0.50 in all cases. Finally, the Variance Inflation Factor (VIF) values remained below 3.3, ruling out the presence of multicollinearity. These results confirm that the instrument applied possesses adequate psychometric properties for measuring the theoretical constructs of the proposed model.
The discriminant validity of the model was then assessed using the Fornell-Larcker criterion, which compares the square root of the AVE with the correlations between the latent variables. As posited by Afthanorhan et al. (2021), the square root of the AVE for each construct must exceed its highest correlation with any other construct in order to demonstrate the discriminant validity of the model. The results, presented in Table 4, confirm that this criterion is met.
Table 4. Fornell-Larcker criterion.
3.2 Analysis of the structure model
The present analysis encompasses both hypothesis testing and the predictive capability of the model. The results of the hypothesis testing are presented in Table 5. In order to conduct this analysis, the path coefficient test was performed, examining each path separately. The findings indicate that all path coefficients are significant at the p = 0.000 level, with additional support from a significance level of p < 0.05 (Astrachan et al., 2014). The findings of the study indicate that seven hypotheses of the model are accepted, while three are rejected.
Table 5. Contrast of structural hypotheses.
The validity of this analysis is contingent upon the model's capacity to provide a comprehensive explanation. The results of this study are presented in Figure 2. The coefficient of determination R2 is utilized to evaluate this power, as it quantifies the predictive value within the sample. According to reference values, an explained variance of 0.72 is considered satisfactory, 0.56 is moderate, and 0.34 is weak (Hair et al., 2017). The findings suggest that behavioral intention and perceived behavioral control possess a moderate-to-high degree of predictive capability. Conversely, perceived usefulness exhibited moderate-low predictive capability, while attitude toward the behavior demonstrated low predictive capacity.
Figure 2. Hypothesis contrast and predictive capacity of the model.
In addition, a non-parametric Stone-Geisser Q2 test was employed to evaluate the extent to which the model approximates expectations. This test establishes values greater than zero as the evaluation criterion (Ringle et al., 2014). However, it is important to note the existence of benchmark values: Predictive accuracy is categorized as low (greater than zero), medium (greater than 0.25), or high (greater than 0.5) (Ghasemy et al., 2020). The findings suggest that perceived behavioral control (Q2 = 0.467) and perceived usefulness (0.412) emerge as the factors exhibiting the greatest predictive capacity.
With respect to demographic characteristics, the sample comprised 105 undergraduate students enrolled in the Public Accounting and Technology in Cost and Budget Analysis programs at the Instituto Tecnológico Metropolitano (ITM) in Medellín, Colombia. Of these, 58.1% were female and 41.9% male. Regarding age demographics, the data revealed that 67.6% of the subjects fell within the 18 to 24 age range, 25.7% were between 25 and 30 years of age, and 6.7% were above 30. With respect to academic progression, 34.3% of the sample were in the first three semesters, 48.6% were in semesters four to six, and 17.1% were in the seventh semester or higher.
While the findings yielded meaningful insights for the ITM student population, it is imperative to acknowledge the limitations of generalizability within the specific institutional and geographical context. The sampling method was based on accessible and voluntary participation, which may introduce a self-selection bias, as reflected in the demographic distribution presented in Table 6. Consequently, the results should be interpreted with caution when attempting to extrapolate them to other institutions, regions, or academic programs. Future research is encouraged to replicate this study using larger and more diverse samples, including both public and private institutions, as well as urban and rural contexts, to enhance the external validity of the findings.
Table 6. Demographic characteristics of the participants.
4 Discussion
The study's findings indicate a robust correlation between perceived ease of use and perceived usefulness. Consequently, students who anticipate minimal technical challenges or complexity when utilizing these tools are more likely to recognize their practical benefits. A substantial body of prior research has identified a positive correlation between these two factors, thereby mediating the relationship between the level of technological readiness exhibited by accounting students and the adoption of artificial intelligence technology (Damerji and Salimi, 2021).
Another significant relationship that has been identified is that of subjective norm versus perceived behavioral control. The study revealed that students' perceptions of social expectations and the influence of prominent individuals served to bolster their confidence in their capacity to utilize AI tools effectively. The findings of this study are corroborated by research conducted by Hwang et al. (2024), which identified a positive correlation between subjective norm and perceived behavioral control in the adoption of artificial intelligence technology.
In a similar vein, subjective norms were also demonstrated to exert influence over students‘ attitudes. This finding suggests that when students perceive that their social environment values and supports the use of this technology, they are more likely to adopt a positive attitude toward its use in accounting education, considering it useful, beneficial, and desirable. The results of the present study underscore the pivotal role of subjective norms in students' technology adoption decisions, thereby facilitating the adoption and approval of specific behaviors (Wang et al., 2024).
The findings further demonstrated that perceived ease of use exerted an influence on perceived behavioral control. Consequently, students' confidence in their ability to use AI tools effectively is increased when they perceive them as easy to handle and understand. The findings of the study demonstrated that control had a significant impact on intention to use. Consequently, students who perceive themselves to possess the necessary skills, resources, and capacity to utilize AI tools are more inclined to express a greater willingness or intention to incorporate them into their learning processes. These results align with the findings of Roy et al. (2022), who determined that perceived behavioral control influences the intention to adopt AI-based robots in university education.
A further salient finding pertains to the influence that subjective norms exert on students' intention to utilize artificial intelligence. In this sense, students are more likely to develop a firm intention to adopt AI in their learning process if they perceive that their social environment values and supports its use. These findings are consistent with those of previous studies that have examined the efficacy of AI-based technologies in the education of university students (Li et al., 2022; Roy et al., 2022).
Conversely, certain hypothesized relationships within the model including those between perceived usefulness and attitude, as well as between perceived usefulness and behavioral intention were determined to be statistically insignificant, as evidenced by p-values that exceeded the 0.05 threshold. However, these results do not indicate a negative influence, since the relationships retain a positive sign. This distinction is of consequence because it suggests that while students may have perceived the tools as useful and easy to use, these perceptions did not have a significant impact on their attitudes or behavioral intentions in this particular context.
The absence of statistically significant relationships may be attributed to several factors. One potential explanation for this phenomenon lies in context-specific expectations or cultural and pedagogical influences that moderate how students translate their perceptions into attitudes or intentions. As Krishnanraw and Ismail (2025) have observed, technology readiness may moderate the relationship between perceived usefulness and attitude, particularly in environments where students have limited prior exposure to AI tools or are accustomed to traditional learning methods. In a similar vein, Bui et al. (2025) found that although Vietnamese accounting students recognized the practicality of AI technologies, their attitudes remained cautious or skeptical due to prevailing educational norms that emphasize conventional over technology-mediated instruction.
This discrepancy may also be indicative of a divergence between the theoretical appreciation and practical applicability of the concept. While students may acknowledge the conceptual benefits of AI tools, they may face challenges in recognizing how these benefits translate into tangible improvements in their own learning outcomes. Al-Okaily (2025) found that, in the context of Chat GPT utilization among accounting students, students' attitudes were more strongly shaped by factors such as trust in the technology, its relevance to the curriculum, and the role played by instructors, despite high levels of perceived usefulness.
In summary, the study revealed that perceived ease of use and perceived usefulness did not have a significant impact on behavioral intention. Additionally, the findings indicated that perceived usefulness did not significantly influence attitude. However, this absence of significance does not necessarily imply a negative influence. Instead, it suggests that other contextual or intervening variables may play a critical role in shaping students‘ acceptance of AI tools. These findings underscore the necessity for strategies that more closely align the perceived usefulness and ease of use of AI tools with students' actual learning experiences and expectations, as well as with institutional and pedagogical support mechanisms.
4.1 Limitations
The limitations of the study are finally highlighted. Firstly, the research was conducted with accounting students from the ITM, which restricts the sample to a single institution. Consequently, the generalizability of the findings may be constrained. It is recommended that subsequent studies encompass an examination of the phenomenon in a variety of higher education institutions, encompassing both urban and rural areas of the country. This approach will facilitate the acquisition of a more comprehensive set of insights and the incorporation of a diverse array of academic and cultural settings.
Secondly, although this study integrates two robust theoretical frameworks the TPB and the TAM the inclusion of other models, such as the UTAUT, may enrich the theoretical foundation. Such inclusion has the potential to provide a more multifactorial and comprehensive perspective on technology adoption, particularly in educational contexts involving emerging technologies such as artificial intelligence.
A notable constraint pertains to the experimental design, which lacked a control group. The absence of a comparative condition imposes limitations on the capacity to isolate the effects of the intervention, specifically the AI-generated video and simulator. In the absence of a baseline or an alternative instructional method, it is challenging to attribute the observed effects exclusively to the intervention. Consequently, this limits the internal validity of the study.
Additionally, the brief exposure to the intervention may have influenced the development of stable perceptions and attitudes. Given that students' interactions with the AI tools were limited to a single session, it is possible that their responses may reflect initial reactions rather than sustained or long-term behavioral intentions. This limited exposure may help explain why some expected relationships such as the influence of perceived usefulness on attitude or intention were found to be statistically insignificant, even though the direction of the associations was positive. The findings of this study underscore the importance of designing longitudinal studies that capture changes in technology acceptance over time and in response to extended engagement with educational technologies.
5 Conclusions
The objective of this study was to examine the behavioral factors that influence accounting students‘ intention to adopt AI-based educational tools, specifically through the use of AI-generated videos combined with interactive simulators. The present study was grounded in the Technology Acceptance Model (TAM) and the Theory of Planned Behavior (TPB). The objective of the present study was to explore the influence of constructs such as perceived usefulness, perceived ease of use, attitude, perceived behavioral control, and subjective norm on students' intention to utilize these tools.
The findings emphasize the pivotal role of perceived ease of use in shaping students' evaluations of the usefulness of AI-based tools in accounting education. When students perceive these tools as intuitive and easy to navigate, they are more likely to recognize their value and relevance to academic learning. This underscores the significance of user-friendly design and interface accessibility in promoting adoption within technology-enhanced learning environments.
Furthermore, the influence of the social environment, particularly the support from peers and instructors, was identified as an important element in enhancing students' confidence in adopting AI tools. However, the findings of this study suggest that such social influences did not demonstrate a statistically significant relationship with behavioral intention. This observation may indicate a need for further exploration of contextual or implementation-related factors.
While a non-significant relationship was identified between perceived usefulness and attitude, the direction of the relationship was not negative. This outcome deviates from conventional expectations within the TAM framework, suggesting that the technology may be in its nascent stages of adoption or that there is a discrepancy between students' expectations and their preliminary experience with the tools. It is evident that further research is necessary to achieve a more profound comprehension of the subtleties inherent to this association.
While the results provide insights into the behavioral mechanisms involved in AI adoption, the conclusions must be interpreted within the limitations of the study. These limitations include the absence of a control group, the short-term exposure to the intervention, and the focus on a single institutional context. Therefore, it is imperative to exercise caution when generalizing the findings.
Future research should replicate and expand these findings across diverse academic and cultural settings, using longitudinal approaches and more robust experimental designs. Furthermore, the integration of alternative or complementary theoretical frameworks has the potential to facilitate a more comprehensive understanding of the multifaceted factors that influence technology adoption in accounting education.
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
Ethical approval was not required for the studies involving humans because the questionnaire was anonymous, with no personal information collected or included. This was previously indicated, and participants would participate in the questionnaire if they accepted the conditions. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants' legal guardians/next of kin in accordance with the national legislation and institutional requirements.
Author contributions
DG: Writing – review & editing, Writing – original draft. AV-A: Writing – original draft, Writing – review & editing. JA: Writing – original draft, Writing – review & editing. JR: Writing – review & editing, Writing – original draft. LV: Writing – review & editing, Writing – original draft. JV: Writing – review & editing, Writing – original draft.
Funding
The author(s) declare that no financial support was received for the research and/or publication of this article.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI Statement
The author(s) declare that no Gen AI was used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
Afthanorhan, A., Ghazali, P. L., and Rashid, N. (2021). Discriminant validity: a comparison of CBSEM and consistent PLS using Fornell and Larcker and HTMT approaches. J. Phys. Conf. Ser. 11874:012085. doi: 10.1088/17422-6596/1874/1/012085
Ajzen, I. (1991). The Theory of planned behavior. Organ. Behav. Hum. Decis. Process. 50, 179–211.doi: 10.1016/07499-5978(91)90020-T
Algerafi, M. A., Zhou, Y., Alfadda, H., and Wijaya, T. T. (2023). Understanding the factors influencing higher education students' intention to adopt artificial intelligence-based robots. IEEE Access 11, 99752–99764. doi: 10.1109/ACCESS.2023.3314499
Alhumaid, K., Naqbi, S., Elsori, D., and Mansoori, M. (2023). The adoption of artificial intelligence applications in education. Int. J. Data Netw. Sci. 7, 457–466. doi: 10.5267/j.ijdns.2022.8.013
Al-Okaily, M. (2024). Artificial intelligence and its applications in the context of accounting and disclosure. J. Financ. Rep. Acc. 8, 435–455. doi: 10.1108/JFRA-04-20244-0209
Al-Okaily, M. (2025). ChatGPT as an educational resource for accounting students: expanding the classical TAM model. Educ. Info. Technol. 8, 435–455. doi: 10.1007/s1s0639-025-133911-1
Al-Okaily, M., Alkhwaldi, A. F., Abdulmuhsin, A. A., Alqudah, H., and Al-Okaily, A. (2023). Cloud-based accounting information systems usage and its impact on Jordanian SMEs' performance: the post-COVID-19 perspective. J. Fin. Rep. Acc. 21, 126–155. doi: 10.1108/JFRA-12-20211-0476
Alotaibi, N. S., and Alshehri, A. H. (2023). Prospers and obstacles in using artificial intelligence in Saudi Arabia higher education institutions—the potential of AI-based learning outcomes. Sustainability 15:10723. doi: 10.3390/su1u51310723
Alshammari, S. H., and Babu, E. (2025). The mediating role of satisfaction in the relationship between perceived usefulness, perceived ease of use and students' behavioural intention to use ChatGPT. Sci. Rep. 15:7169. doi: 10.1038/s4s1598-025-916344-4
Amoako-Gyampah, K. (2007). Perceived usefulness, user involvement and behavioral intention: an empirical study of ERP implementation. Comput. Hum. Behav. 23, 1232–1248. doi: 10.1016/j.chb.2004.12.002
Aschwanden, D., Strickhouser, J. E., Sesker, A. A., Lee, J. H., Luchetti, M., Terracciano, A., et al. (2021). Preventive behaviors during the COVID-19 pandemic: associations with perceived behavioral control, attitudes, and subjective norm. Front. Public Health 9:662835. doi: 10.3389/fpubh.2021.662835
Astrachan, C. B., Patel, V. K., and Wanzenried, G. (2014). A comparative study of CB-SEM and PLS-SEM for theory development in family firm research. J. Fam. Bus. Strateg. 5, 116–128. doi: 10.1016/j.jfbs.2013.12.002
Baba, R., Keling, W., and Yap, C. S. (2025). The effect of subjective norms, attitude and start-up capital on the entrepreneurial intention of the indigenous people in Malaysia. J. Enterp. Commun. Ppl. Pl. Glob. Econ. 9, 853–880. doi: 10.1108/JEC-05-20244-0100
Bai, X., Guo, R., and Gu, X. (2024). Effect of teachers' TPACK on their behavioral intention to use technology: chain mediating effect of technology self-efficacy and attitude toward use. Educ. Inform. Technol. 29, 1013–1032. doi: 10.1007/s1s0639-023-12343-x
Bui, H. Q., Phan, Q. T. B., and Nguyen, H. T. (2025). AI adoption: a new perspective from accounting students in Vietnam. J. Asian Bus. Econ. Stud. 32, 40–51. doi: 10.1108/JABES-06-20244-0300
Cao, A., Guo, L., and Li, H. (2025). Understanding farmer cooperatives' intention to adopt digital technology: mediating effect of perceived ease of use and moderating effects of internet usage and training. Int. J. Agric. Sustain. 23:2464523. doi: 10.1080/14735903.2025.2464523
Damerji, H., and Salimi, A. (2021). Mediating effect of use perceptions on technology readiness and adoption of artificial intelligence in accounting. Acc. Educ. 30, 107–130. doi: 10.1080/09639284.2021.1872035
Davis, F. D., Bagozzi, R. P., and Warshaw, P. R. (1989). Technology acceptance model. J. Manag. Sci. 35, 982–1003. doi: 10.1287/mnsc.35.8.982
Denovan, R. F., and Marsasi, E. G. (2025). Perceived ease of use, perceived usefulness and satisfaction to maximize behavioral intention with the technology acceptance model in generation y and z consumers. J. Pamator 18, 1–36. doi: 10.21107/pamator.v1v8i88i1ii1.29461
Essam, E., Abdien, M. K., Barakat, M. O., and Elsaqqa, M. A. (2025). Unveiling locavorism in restaurants: understanding consumer's intention and the dual moderating roles of subjective norms and perceived behavioral control. J. Hosp. Tour. Insights 8, 2232–2252. doi: 10.1108/JHTI-07-20244-0692
Gaviria, D., Arango, J., Valencia-Arias, A., Palacios-Moya, L., Velez Holguin, R., and Gallegos Ruiz, A. L. (2022). Factors that affect the usage intention of virtual learning objects by college students. Informatics 9:65. doi: 10.3390/informatics9s030065
Ghasemy, M., Teeroovengadum, V., Becker, J. M., and Ringle, C. M. (2020). This fast car can move faster: a review of PLS-SEM application in higher education research. High. Educ. 80, 1121–1152. doi: 10.1007/s1s0734-020-005344-1
Gruenbichler, R., Wijaya, L., Meng, C. K., Greimel, K., Anita, T. L., and Samuel, S. (2024). “Artificial intelligence adoption among accountants: empirical study in Austria,” in 2024 International Conference on Information Management and Technology (ICIMTech) (New York, NY: IEEE), 281–286. doi: 10.1109/ICIMTech6h3123.2024.10780935
Hair, J. F., Matthews, L. M., Matthews, R. L., and Sarstedt, M. (2017). PLS-SEM or CB-SEM: updated guidelines on which method to use. Int. J. Multivar Data Anal. 1, 107–123. doi: 10.1504/IJMDA.2017.087624
Hair, J. F., Ringle, C. M., and Sarstedt, M. (2011). PLS-SEM: indeed a Silver Bullet. J. Market. Theor. Pract. 19, 139–152. doi: 10.2753/MTP1P069-6679190202
Hair, J. F., Risher, J. J., Sarstedt, M., and Ringle, C. M. (2019). When to use and how to report the results of PLS-SEM. Eur. Bus. Rev. 31, 2–24. doi: 10.1108/EBR-11-20188-0203
Ho, J. C., Wu, C. G., Lee, C. S., and Pham, T. T. T. (2020). Factors affecting to adopt mobile banking: an international comparison. Technol. Soc. 63:101360. doi: 10.1016/j.techsoc.2020.101360
Hoque, M. E., Susanto, P., Shah, N. U., Khatimah, H., and Mamun, A. A. (2024). Does perceived behavioral control mediate customers' innovativeness and continuance intention of e-money? The moderating role of perceived risk and e-security. Int. J. Emerg. Mark. 19, 4481–4502. doi: 10.1108/IJOEM-06-20222-0914
Humida, T., Al Mamun, M. H., and Keikhosrokiani, P. (2022). Predicting behavioral intention to use e-learning system: a case-study in Begum Rokeya University, Rangpur, Bangladesh. Educ. Inform. Technol. 27, 2241–2265. doi: 10.1007/s1s0639-021-107077-9
Hwang, J., Kim, J., Joo, K. H., and Choe, J. Y. (2024). An integrated model of artificially intelligent (AI) facial recognition technology adoption based on perceived risk theory and extended TPB: a comparative analysis of US and South Korea. J. Hosp. Market. Manag. 33, 1071–1099. doi: 10.1080/19368623.2024.2379269
Kaba, A., Eletter, S., Ramaiah, C. K., and El Refae, G. A. (2025). Demographic differences in attitude, subjective norms, behavioral intention, and knowledge sharing behavior: an empirical study of non-academic staff from India and the UAE. VINE J. Inform. Knowl. Manag. Syst. 55, 470–491. doi: 10.1108/VJIKMS-07-20222-0235
Kim, Y. W., Cha, M. C., Yoon, S. H., and Lee, S. C. (2025). Not merely useful but also amusing: impact of perceived usefulness and perceived enjoyment on the adoption of AI-powered coding assistant. Int. J. Hum. Comput. Int. 41, 6210–6222. doi: 10.1080/10447318.2024.2375701
Krishnanraw, J., and Ismail, K. (2025). Behavioural intention to use artificial intelligence (AI) among accounting students: evaluating the effect of technology readiness. Manag. Acc. Rev. (MAR) 24, 465–493. doi: 10.24191/MAR.V2V4i44i0i1-17
La Barbera, F., and Ajzen, I. (2020). Control interactions in the theory of planned behavior: rethinking the role of subjective norm. Eur. J. Psychol. 16:401. doi: 10.5964/ejop.v1v6i66i3ii3.2056
Li, X., Jiang, M. Y., Jong, M. S., Zhang, X., and Chai, C. (2022). Understanding medical students' perceptions of and behavioral intentions toward learning artificial intelligence: a survey study. Int. J. Environ. Res. Public Health 19:8733. doi: 10.3390/ijerph1h9148733
Margraf, J., Brailovskaia, J., and Schneider, S. (2020). Behavioral measures to fight COVID-19: an 8-country study of perceived usefulness, adherence and their predictors. PLoS ONE 15:e0e243523. doi: 10.1371/journal.pone.0243523
Martínez Ávila, M., and Fierro Moreno, E. (2018). Aplicación de la técnica PLS-SEM en la gestión del conocimiento: un enfoque técnico práctico/Application of the PLS-SEM technique in Knowledge Management: a practical technical approach. RIDE 8, 130–164. doi: 10.23913/ride.v8vi88i1i6.336
Norzelan, N. A., Mohamed, I. S., and Mohamad, M. (2024). Technology acceptance of Artificial Intelligence (AI) among heads of finance and accounting units in the shared service industry. Technol. Forecast. Soc. Change 198:123022. doi: 10.1016/j.techfore.2023.123022
Oematan, M. E., Rahayu, S., and Trisnawati, J. D. (2024). The effect of perceived usefulness and perceived ease of use on behavioral intention mediated by user satisfaction in e-commerce users. J. Ekon. 13, 472–479. Available online at: https://ejournal.seaninstitute.or.id/index.php/Ekonomi/article/view/4914
Pillai, R., Sivathanu, B., Metri, B., and Kaushik, N. (2024). Students' adoption of AI-based teacher-bots (T-bots) for learning in higher education. Inform. Technol. People 37, 328–355. doi: 10.1108/ITP-02-20211-0152
Purwanto, A., and Sudargini, Y. (2021). Partial Least Squares Structural Squation Modeling (PLS-SEM) analysis for social and management research : a literature review. J. Indust. Engin. Manag. Res. 2, 114–123. doi: 10.7777/jiemar.v2i4.168
Rasoolimanesh, S. M. (2022). Discriminant validity assessment in PLS-SEM: a comprehensive composite-based approach. Data Anal. Perspect. J. 3, 1–8.
Ringle, C., Wende, S., and Becker, J. (2022). SmartPLS 4. Bonningstedt: SmartPLS GmbH. Available online at: http://www.smartpls.com (Accessed November 7, 2024).
Ringle, C. M., Da Silva, D., and Bido, D. D. S. (2014). Modelagem de Equações Estruturais com Utilização do Smartpls. Rev. Bras. Mark. 13, 56–73. doi: 10.5585/remark.v1v3i33i2ii2.2717
Roy, R., Babakerkhell, M. D., Mukherjee, S., Pal, D., and Funilkul, S. (2022). Evaluating the intention for the adoption of artificial intelligence-based robots in the university to educate the students. IEEE Access 10, 125666–125678. doi: 10.1109/ACCESS.2022.3225555
Sudaryanto, M. R., Hendrawan, M. A., and Andrian, T. (2023). “The effect of technology readiness, digital competence, perceived usefulness, and ease of use on accounting students Artificial Intelligence Technology Adoption,” in E3ES33S Web of Conferences (EDP Sciences), doi: 10.1051/e3es33sconf/202338804055
Tahar, A., Riyadh, H. A., Sofyani, H., and Purnomo, W. E. (2020). Perceived ease of use, perceived usefulness, perceived security and intention to use e-filing: the role of technology readiness. J. Asian Fin. Econ. Bus. 7, 537–547. doi: 10.13106/jafeb.2020.vol7ll7.no9oo9.537
Valencia-Arias, A., Gómez-Molina, S., Vélez-Holguín, R. M., and Cardona-Acevedo, S. (2023). Intención de uso de aprendizaje móvil (m-learning) en programas virtuales: un modelo híbrido de aceptación tecnológica (TAM) y la teoría del comportamiento planificado (TPB). Form. Univ. 16, 25–34. doi: 10.4067/S0S718-50062023000200025
Vamvaka, V., Stoforos, C., Palaskas, T., and Botsaris, C. (2020). Attitude toward entrepreneurship, perceived behavioral control, and entrepreneurial intention: dimensionality, structural relationships, and gender differences. J. Innov. Entrepreneurship 9, 1–26. doi: 10.1186/s1s3731-020-01122-0
Wang, C., Wang, H., Li, Y., Dai, J., Gu, X., and Yu, T. (2024). Factors influencing university students' behavioral intention to use generative artificial intelligence: integrating the theory of planned behavior and AI literacy. Int. J. Hum. Comput. Interact. 41, 6649–6671. doi: 10.1080/10447318.2024.2383033
Warsono, H., Yuwono, T., and Putranti, I. R. (2023). Analyzing technology acceptance model for collaborative governance in public administration: empirical evidence of digital governance and perceived ease of use. Int. J. Data Netw. Sci. 7, 41–48. doi: 10.5267/j.ijdns.2022.12.008
Wiprayoga, P., Gede, S., and Suasana, G. A. K. G. (2023). The role of attitude toward using mediates the influence of perceived usefulness and perceived ease of use on behavioral intention to use. Russ. J. Agric. Soc. Econ. Sci. 140, 53–68. doi: 10.18551/rjoas.20233-08.06
Wismantoro, Y., and Susilowati, M. W. K. (2025). Do attitude toward behavior, subjective norms, and perceived control behavior matter on environmentally friendly plastic purchasing intention?. Int. J. Manag. Sustain. 14, 44–56. doi: 10.18488/11.v1v4i44i1ii1.4000
Wu, H., Wang, Y., and Wang, Y. (2024). “To Use or Not to Use?” A mixed-methods study on the determinants of EFL college learners' behavioral intention to use AI in the distributed learning context. Int. Rev. Res. Open Distrib. Learn. 25, 158–178. doi: 10.19173/irrodl.v2v5i55i3ii3.7708
Zeng, N., Liu, Y., Gong, P., Hertogh, M., and König, M. (2021). Do right PLS and do PLS right: a critical review of the application of PLS-SEM in construction management research. Front. Engin. Manag. 8, 356–369. doi: 10.1007/s4s2524-021-01533-5
Keywords: perceived usefulness, perceived ease of use, behavioral intention, higher education, subjective norm, perceived behavioral control
Citation: Gaviria Rodríguez DY, Valencia-Arias A, Arango Arango JG, Raunelli Sander JM, Verde Flores L and Valencia J (2025) Intention to use AI in accounting education: an analysis from the TAM and TPB perspectives. Front. Educ. 10:1637857. doi: 10.3389/feduc.2025.1637857
Received: 29 May 2025; Accepted: 06 August 2025;
Published: 02 September 2025.
Edited by:
Chayanika Uniyal, University of Delhi, IndiaReviewed by:
Rajat Sant, University of Delhi, IndiaRasha Nouraldeen, Beirut Arab University, Lebanon
Copyright © 2025 Gaviria Rodríguez, Valencia-Arias, Arango Arango, Raunelli Sander, Verde Flores and Valencia. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Alejandro Valencia-Arias, dmFsZW5jaWFqaG9AdXNzLmVkdS5wZQ==
†ORCID: Diana Yanet Gaviria Rodríguez orcid.org/0000-0002-4525-1423
Alejandro Valencia-Arias orcid.org/0000-0001-9434-6923
Juan Guillermo Arango Arango orcid.org/0000-0002-0084-7362
Juan Manuel Raunelli Sander orcid.org/0000-0001-5818-949X
Lilian Verde Flores orcid.org/0000-0003-0496-853X
Jackeline Valencia orcid.org/0000-0001-6524-9577