Possessing a belief entail possessing a perceived truth, which subsequently dictates the subject’s actions and behavior in both advantageous and disadvantageous manners. The rapid pace of cultural, social, and technological transformation necessitates a reevaluation of personal beliefs, much like values, within the dynamic social reality we inhabit.

Beliefs, in their most expansive sense, wield transformative power over our cognition, opinions, competencies, and attitudes. They extend beyond mere personal constructs and are significantly influenced by the broader cultural milieu. As Jiménez (2019, p. 16) aptly articulates, “Beliefs are general ideas that we have about phenomena; they help us interpret, predict and control the events that occur and thus make decisions.” This conceptualization of beliefs is paramount in understanding their role in shaping our actions and behaviors, particularly for educators who must recognize that their pedagogical choices are profoundly influenced by their beliefs.

For a considerable duration (Brody and Day, 1993), it has been posited that in the realm of teaching, the significance and value that educators ascribe to their beliefs delineate the meaning of education and its various components (such as content, technology, didactic strategies, evaluation methods, etc.), and consequently influence their conduct in the educational process, their perception of it, and the nature of their interactions with students. Clark and Peterson (1986), pioneers in this research domain, assert that teachers’ cognitive frameworks impact their perceptions of students, their instructional planning, and their classroom behavior.

Montanares and Junod (2018, p. 94) further propose that “knowledge of teachers’ beliefs is important to the extent that it allows teachers a greater degree of awareness, responsibility and control in the choices of epistemological teaching models.”

The formation of teachers’ beliefs is influenced by a diverse array of variables, encompassing their professional experiences as educators and researchers, the nature of their training, their pedagogical education, ideological inclinations, interactions with colleagues, and the characteristics and ideology of the educational institution and community in which they operate (Arancibia-Herrera et al., 2024; Palacios-Rodríguez et al., 2023).

From the perspective of didactic research, investigations into beliefs have pursued various lines and orientations, including teaching strategies (Montanares and Junod, 2018; Cabero-Almenara et al., 2024), instructional practices (Rodríguez-Sosa and Solis-Manrique, 2017; Al-Adwan & Al-Debei, 2024), discipline-specific pedagogy (Casimiro, 2022), inclusive education (Friesen et al., 2023), the pedagogical utility of questioning (Mante-Estacioa and Tupas, 2023), STEM education (Alghamdi, 2023), and the role of emotions in professional teaching practices (Vettori et al., 2022).

A significant research trajectory has also examined beliefs about educational possibilities and the integration of Information and Communication Technologies (ICT). This research has evolved to focus on specific analyses, such as perceptions regarding the TPACK model’s dimensions in ICT integration (Ifinedo et al., 2020), the efficacy of video in language instruction (Waluyo and Apridayani, 2021), and attitudes toward incorporating ICT into educational practice (Li et al., 2019; Hoareau et al., 2021). Additionally, studies have explored how beliefs about technological competencies influence ICT integration (Cheng et al., 2022), the potential of ICT to support students with dyslexia (Bice and Tang, 2022), its effectiveness in lower education levels (Hoareau et al., 2021), and its applicability across various disciplines (García et al., 2022).

To summarize the significance of teachers’ beliefs concerning ICT application, Tondeur et al. (2017) conducted a meta-analysis and concluded: (1) a bidirectional relationship exists between pedagogical beliefs and ICT usage, (2) beliefs often act as perceived barriers, (3) specific beliefs are linked to specific types of ICT usage, (4) beliefs play a crucial role in professional development, and (5) the school context significantly influences beliefs about ICT.

Related to the theme of beliefs is the work on the “degree of acceptance of technologies” by teachers. Various models have been employed to analyze technology acceptance among potential users. The initial model, the Technology Acceptance Model (TAM) formulated by Davis (1989), posits that the intention to use technology is influenced by two primary dimensions: perceived usefulness and perceived ease of use, which in turn affect attitudes toward ICT, determining intentions to use and actual usage. This model has been utilized to analyze different technologies, such as virtual training (Rodríguez-Sabiote et al., 2023), augmented reality (Barroso et al., 2018), and immersive reality (Cabero et al., 2023; Palacios-Rodríguez et al., 2024a).

In contrast to the Technology Acceptance Model (TAM), Venkatesh et al. (2003) synthesized various proposed acceptance models, including TAM, to develop the Unified Theory of Acceptance and Use of Technology (UTAUT). This model aims to elucidate the acceptance and utilization of technology, predicated on four primary dimensions: performance expectancy, effort expectancy, social influence, and facilitating conditions. Subsequently, Venkatesh et al. (2012) refined the model, introducing the UTAUT2, which incorporates three additional dimensions: hedonic motivation (the pleasure derived from using the technology), price value, and the degree of automatic technology use. This reformulated model has garnered increasing adoption among researchers, as noted by Tamilmani et al. (2021) and García et al. (2022), compared to its predecessors.

It is noteworthy that our study exclusively considers the first of the new variables introduced in UTAUT2—hedonic motivation. The variables of price value and automatic technology use were deemed irrelevant for the specific objectives of our research, thus configuring the model as depicted in Figure 1.

Understanding, according to various studies (Gansser and Reich, 2021; Al-Adwan and Al-Debei, 2024; Marikyan and Papagiannidis, 2023), the following dimensions are articulated within the UTAUT and UTAUT2 frameworks:

It should be noted that the UTAUT and UTAUT2 models have been widely employed to assess the acceptance of various technologies. Within the UTAUT framework, research has explored the acceptance of the metaverse (Lee and Kim, 2022), mobile devices (Mojarro et al., 2019), and virtual reality (Ustun et al., 2023). Meanwhile, UTAUT2 has been applied to analyze the acceptance of technologies such as augmented reality (Huang, 2020), the metaverse (Al-Adwan and Al-Debei, 2024), virtual training platforms (Zacharis and Nikolopoulou, 2022), artificial intelligence (Gansser and Reich, 2021), and its educational applications (Strzelecki, 2023).

In short, these models enable us to comprehend whether teachers are inclined to use Educational Artificial Intelligence (AIEd), the extent of its usage, and the factors influencing these decisions, such as peer opinions, prior experiences, perceived value, and ease of use.

The recent advancements in Artificial Intelligence, particularly since the introduction of ChatGPT-3, have significantly impacted educational research. This progress has spurred investigations into attitudes, acceptance levels, necessary training, and the role of teachers’ beliefs in shaping students’ utilization of AI in educational settings. The importance of training educators in AI for both pedagogical and research purposes has also been highlighted (Alenezi et al., 2023; Tongfei, 2023; González-Mayorga et al., 2024; Temitayo et al., 2024).

Teacher beliefs are influenced by age, as demonstrated by Yuk and Lee (2023), who explored the experiences, perceptions, knowledge, concerns, and intentions of Generation Z students compared to older generations of educators. Their findings indicate generational differences in the adoption of technology in teaching.

Regarding AI, Adekunle et al. (2022) found that teachers’ confidence in teaching AI predicts their intention to incorporate AI into their instruction, suggesting that educators’ beliefs about AI’s usefulness and educational relevance are crucial. However, these beliefs are not uniform and vary depending on the discipline and educational level (Delgado et al., 2024).

From a psychoeducational perspective, two contrasting positions regarding conceptions of learning and teaching are prevalent: behaviorist and constructivist perspectives. The behaviorist approach suggests that knowledge is transmitted to the learner, while the constructivist perspective posits that knowledge is constructed subjectively and through social interaction (Arancibia, 2022). Choi et al. (2023) found that teachers with constructivist beliefs are more likely to integrate AI into their teaching than those with transmissive orientations.

Furthermore, studies have shown that teachers’ beliefs about learning impact their use of ICT in teaching and the frequency of its use. This includes research on general ICT usage (Prestridge, 2017; Bahçivan et al., 2018; Li et al., 2019; Arancibia-Herrera et al., 2024; Palacios-Rodríguez et al., 2024a, 2024b) and specific technologies such as mixed reality (Marín et al., 2023), the Moodle platform (Arancibia, 2022), and digital whiteboards (Burke et al., 2018).

The conclusions of our study point in different directions. Firstly, the instrument’s reliability has been demonstrated to exhibit reasonably high levels when analyzing the dimensions identified from the UTAUT2 model and the constructivist and transmissive pedagogical perspectives. These values align with those found by Choi et al. (2023) and Cabero-Almenara et al. (2024).

Simultaneously, the model formulated to analyze the degree of teachers’ acceptance of AIEd proved reliable and valid. The results confirmed the significance of 16 of the 17 formulated hypotheses, with the exception of the hypothesis relating to Effort Expectancy (EE) in Attitude Toward Use (ATT). This discrepancy could be attributed to the ease of using generative AI tools and the lack of knowledge or training among teachers. Future research should consider directly addressing questions about the degree of technical and didactic mastery specifically related to AIEd, rather than to technologies in general.

It is essential to highlight that Attitude Toward Use (ATT) is the most significant and influential dimension concerning Behavioral Intention (BI). Therefore, the intention to use fundamentally determines and directs the actual use by the teacher. Moreover, the results indicate that teachers with constructivist beliefs are more likely to integrate AIEd into teaching than those with transmissive orientations. This finding corroborates the results achieved by Choi et al. (2023). However, in this study, the acceptance model used was the UTAUT2. Only in the case of Effort Expectancy (EE) did teachers with transmissive beliefs establish a higher score load, potentially indicating their perception of greater difficulty in using AIEd.

The findings of this study have both practical and theoretical implications for understanding AIEd. To our knowledge, this is one of the first empirical studies to address teachers’ perceptions of AIEd through the UTAUT2 technology acceptance model. Previous studies have utilized the TAM model of technology acceptance (Choi et al., 2023) or the UTAUT2 model with university students (Strzelecki, 2023).

This study developed a conceptual model derived from the UTAUT2 and integrated the potential significance of teachers’ transmissive or constructivist pedagogical beliefs for the use of AIEd. Through a structural equation modeling (SEM) analysis, it is posited that the proposed model offers an adequate explanation of teachers’ intentions to use AIEd. This study has highlighted the importance of considering Performance Expectancy (PE), Effort Expectancy (EE), Social Influence (SI), Facilitating Conditions (FC), and Hedonic Motivation (HM) in predicting the acceptance of AIEd and its impact on attitudes and behavioral intentions to use (Al-Adwan & Al-Debei, 2024).

The findings reveal that teachers with a more constructivist approach positively correlate with the perceived usefulness and effectiveness of AIEd, leading to a higher likelihood of adoption. Consequently, educational institutions must promote a constructivist-oriented pedagogical perspective while providing teacher training programs to familiarize them with AIEd and enhance their digital competencies for integration into the educational environment.

However, this research has limitations that should be addressed in future studies. Firstly, the extent to which respondents were exposed to AIEd was not considered. Additionally, the demographic characteristics of the teachers, such as age, gender, and subjects taught, were not included in the analyses but will be addressed in future work currently in preparation.

Finally, incorporating qualitative data would provide a more comprehensive investigation into the determinants influencing teachers’ acceptance of AIEd. Combining qualitative methods, such as focus group interviews or nominal group techniques, could enable researchers to unravel the underlying mechanisms involved in this acceptance process.