Exploring Saudi EFL Learners’ Engagement with ChatGPT: A Mixed-Methods Study of Perceptions,Attitudes,and Intentions

Abstract

This study explores Saudi EFL learners’ perceptions, attitudes, and intentions regarding using ChatGPT as an AI-mediated tool for English language learning. Employing a mixed-methods design, 181 students completed a survey, and 4 participated in follow-up interviews to offer a more nuanced understanding of learner experiences. Quantitative findings reveal high perceived usefulness (M = 3.9), a generally pleasant user experience (M = 3.95), and a firm intention to continue using the tool (M = 3.7). Pearson correlation analysis confirmed significant positive relationships between perception and intention (r = .797, p < .001) and between attitude and intention (r = .733, p < .001), thereby reinforcing the explanatory power of the Technology Acceptance Model (TAM) in the Saudi EFL context. Qualitative findings add depth to the statistical trends, showcasing that Saudi EFL considered ChatGPT not merely a functional platform but an emotionally supportive and context-sensitive mediator in their academic language development. Students demonstrated critical agency, using the tool strategically for scaffolding, meaning-making, and managing learning anxiety while remaining vigilant about its limitations. Participants expressed concerns about over-reliance on AI, warning against diminished critical thinking, intellectual complacency, and the erosion of human interaction in education. Framed through the dual lenses of TAM and sociocultural theory, the study argues that learners do not passively adopt ChatGPT; instead, they negotiate its role as a complementary and temporary scaffold that enhances but does not replace human faculties and social learning environments. Thus, the study underscores the need for pedagogically grounded and ethically informed integration of AI into language education. These findings offer valuable implications for designing more balanced, culturally responsive, and learner-centered approaches to technology use in Saudi EFL classrooms.

Plain Language Summary

Exploring Saudi EFL Learners’ Engagement with ChatGPT: A Mixed-Methods Study of Perceptions, Attitudes, and Intentions

This study explores Saudi EFL learners’ perceptions, attitudes, and intentions regarding using ChatGPT as an AI-mediated tool for English language learning. Employing a mixed-methods design, 181 students completed a survey, and four participated in follow-up interviews to offer a more nuanced understanding of learner experiences. Quantitative findings reveal high perceived usefulness (M = 3.9), a generally pleasant user experience (M = 3.95), and a firm intention to continue using the tool (M = 3.7). Pearson correlation analysis confirmed significant positive relationships between perception and intention (r = .797, p < .001) and between attitude and intention (r = .733, p < .001), thereby reinforcing the explanatory power of the Technology Acceptance Model (TAM) in the Saudi EFL context. Qualitative findings add depth to the statistical trends, showcasing that Saudi EFL considered ChatGPT not merely a functional platform but an emotionally supportive and context-sensitive mediator in their academic language development. Students demonstrated critical agency, using the tool strategically for scaffolding, meaning-making, and managing learning anxiety while remaining vigilant about its limitations. Participants expressed concerns about over-reliance on AI, warning against diminished critical thinking, intellectual complacency, and the erosion of human interaction in education. Framed through the dual lenses of TAM and sociocultural theory, the study argues that learners do not passively adopt ChatGPT; instead, they negotiate its role as a complementary and temporary scaffold that enhances but does not replace human faculties and social learning environments. Thus, the study underscores the need for pedagogically grounded and ethically informed integration of AI into language education.

Keywords

educational research education social sciences distance learning higher education applied linguistics linguistics language studies humanities language teaching education theory and practice

Introduction

In the contemporary landscape of education, technology has become an integral and transformative force, reshaping traditional teaching methods and redefining the learning experience (Aljabr & Al-Ahdal, 2024; Rospigliosi, 2023). The ubiquitous presence of technology in classrooms and beyond has brought about significant transformations, fostering innovation, accessibility, autonomy, and engagement (Adeshola & Adepoju, 2023). Along these lines of thoughts, the emerging dynamics of the education and research paradigm explore the multifaceted role of technology in modern education, examining its impact on teaching approaches, student learning experiences, and the broader educational landscape (Tuomi, 2019).

Information is now not the prerogative of a few; rather, there is enhanced access to it like never before. Technology has increased equitable access to information, breaking geographical barriers and providing learners with knowledge (Alowedi & Al-Ahdal, 2023). The internet, digital libraries, and online resources empower students to explore diverse perspectives and deepen their understanding of various subjects. The democratization of information has levelled the playing field, ensuring that education is no longer confined to traditional brick-and-mortar institutions, nor is it merely a banking system where students are only receivers of knowledge (Oliver, 2008; Rospigliosi, 2022). Opportunities for personalized learning have become a reality with adaptive learning platforms and intelligent tutoring systems leveraging technology to cater to individuals with unique learning styles, foster agency, and adjust pacing (Taylor et al., 2021). With the ability to tailor content and assessments based on student performance, AI-mediated tools enable a personalized learning experience (Strzelecki, 2023). This individualization fosters a more inclusive and empowering educational environment, accommodating diverse learning needs, and maximizing student potential (Holmes et al., 2018). In today’s ever-fast-paced learning opportunities, AI-mediated platforms have transformed pedagogical practices (Benuyenah, 2023). These tools make lessons and course content more engaging and interactive to capture students’ attention and promote active participation and agency (Malik et al., 2023). In English language learning, for instance, AI-assisted tools contribute to immersive learning experiences, making complex concepts more accessible and stimulating students’ curiosity (Sviridova et al., 2023).

In the context of English as a foreign language (EFL), the capabilities of AI-generated text and the types of writing produced by applying AI functions—such as translating, co-thinking, editing, paraphrasing, revising, and text modification (Barrot, 2023)—demonstrate the great potential ChatGPT offers for EFL students. However, it also presents significant risks and challenges (Kolade et al., 2024). From this perspective, it is crucial to explore the experiences of university language students regarding their perceived usage of ChatGPT and the potential challenges of integrating it into their learning journey (Barrot, 2023; Strzelecki, 2023). The readiness of students to adopt technology in education is a dynamic, complex, and multifaceted process (Strzelecki, 2023), requiring a concerted effort from educational institutions, policymakers, and the broader community to address issues related to digital literacy, access, and attitudes (Wohlfart & Wagner, 2023).

The cultural characteristics of the Saudi educational context are more traditional teacher and often culture-based collective societies, with students being dependent and using others as a source of support (Alhammad, 2024; Alharthi, 2024). Nonetheless, there is a lack of research that examines the extent to which Saudi EFL students can use ChatGPT effectively, instead of a superficial evaluation of its utility. The mixed-methods design offers a context-sensitive and theoretically grounded account—drawing on the Technology Acceptance Model (TAM)—of learners’ perceptions, attitudes, and behaviors, with a view toward the future implications of AI tools in language learning. Specific attention is given to the sociocultural dimensions of local students’ identities (Lantolf & Thorne, 2006; Vygotsky, 1978) and the development of learner agency (Benson, 2013; Miller, 2012) within an AI-mediated environment that supports co-thinking and languaging in language learning (Swain & Lapkin, 2013).

Literature Review

Elbanna and Armstrong (2024) and Hutson et al. (2022) state that OpenAI’s ChatGPT, an advanced AI-driven language learning model, has gained distinction by surpassing 100 million users, which continues to grow rapidly. There is increasing interest in its adoption, particularly within higher education institutions. Using OpenAI’s ChatGPT has shown notable growth in domestic and international academic contexts. Selected studies exploring its educational application have been reviewed in this section, highlighting key findings and research trends related to its impact, learner attitudes, and user experiences. What follows are the global and local empirical findings on integrating AI tools in education and language learning contexts.

Global Perspectives on ChatGPT in Education

Recent concerns about AI replacing human intervention, particularly in education, have prompted researchers to explore AI tools like ChatGPT’s potential benefits and challenges. For example, Acosta-Enriquez et al. (2024) found that tool acceptance, frequency of use, and positive emotions influenced attitudes toward ChatGPT among adult learners. However, users also tended to verify information obtained from ChatGPT, indicating that reliance on human agency is critical over chatbot-generated content. Similarly, Abdaljaleel et al. (2024) identified demographic factors such as country of residence, age, university type, and academic performance as significant predictors of students’ attitudes toward ChatGPT usage. These studies show that learners are not unquestioningly accepting AI; their perceptions are mediated by emotional, demographic, and contextual variables, challenging the techno-optimistic tone of early adoption research.

Furthermore, Yu et al. (2024) highlighted ChatGPT’s dual role as both a practical tool and an emotional support system for students, noting concerns about content falsification and job displacement. In a more in-depth analysis of ChatGPT usability, Raman et al. (2023) utilized Rogers’ (2003) perceived theory of qualities to investigate ChatGPT adoption among 288 university students. The study identified five key characteristics, relative advantage, compatibility, ease of use, observability, and trialability, that significantly influenced adoption. Male students prioritized reliability, straightforward use, and visibility, while female students emphasized reliability, ease of use, comparative advantage, and feasibility tests. While this research advances adoption theory, it lacks engagement with pedagogical nuance, particularly around how AI tools intersect with literacy development and language learning practices.

Exploring agency, Agustini (2023) showcased ChatGPT’s role in fostering independence and agency among first-year college students in the Kurikulum Merdeka Belajar, particularly in English language learning. The study highlighted ChatGPT’s potential to provide individualized support, language practice opportunities, self-reflection, self-assessment, and instant feedback, promoting student autonomy and self-directed learning. However, such promising accounts often overlook learners’ critical awareness, ethical reservations, or institutional constraints.

Similarly, Haque et al. (2022) analyzed 10,732 tweets from early ChatGPT users, identifying positive sentiments related to entertainment, creativity, and software development. Concerns about ChatGPT’s impact on education were less prominent, though the study emphasized the importance of user sentiment in determining the success of new technologies. Expanding on this, Taecharungroj (2023) analyzed 233,914 English tweets using latent Dirichlet allocation (LDA) topic modeling, identifying three main themes: news, technology, and reactions, and five functional domains: creative writing, essay writing, prompt writing, code writing, and question answering. The findings underscored ChatGPT’s potential to influence technology and society positively and negatively, raising critical issues such as job evolution, artificial general intelligence, and ethical standards.

Collectively, these global studies highlight diverse uses of ChatGPT, but most lack specific inquiry into how AI functions within language learning ecosystems marked by exam-driven curricula and sociocultural sensitivities. At a broader level, Perkins (2023) examined the ethical implications of students using ChatGPT and other AI tools in formal exams. The study highlighted the potential contributions of large language models (LLMs) to teaching, human-AI collaboration, and automated writing evaluations. However, it also raised concerns about academic integrity, particularly the ability of AI tools to produce undetectable, original text. Perkins argued that institutional policies, rather than student use, are crucial in maintaining academic integrity in the face of advancements in the AI era. This concern is echoed in Germany, where Moldt et al. (2023) found that medical students lacked trust in ChatGPT due to data protection concerns, particularly in sensitive disciplines like medicine. These studies collectively signal growing ethical unease about AI tools. Nevertheless, few offer insight into how learners feel and navigate these tensions in EFL environments, where norms of authorship and originality are still evolving.

Local and Regional Perspectives

In Saudi Arabia, Alharthi (2024) surveyed 299 undergraduates across 3 private universities, where students’ English proficiency is considerably high and their exposure to tech-based learning tools is relatively frequent. The study revealed generally positive attitudes toward ChatGPT as a supportive, accessible, and user-friendly tool. Nonetheless, concerns about overreliance and its potential impact on students’ critical thinking and independent writing abilities were raised. Complementary studies by Alhammad (2024) and Masoudi (2024) also highlighted ChatGPT’s role in enhancing literature comprehension and writing proficiency, particularly by assisting with drafting and revision tasks. Building on this line of inquiry, Alotaibi et al. (2025) employed the Technology Acceptance Model (TAM) to examine Saudi EFL learners’ behavioral intentions. Their findings confirmed the importance of perceived usefulness, ease of use, and attitude in predicting ChatGPT adoption, especially for grammar assistance, vocabulary support, and idea generation. While valuable, this study primarily emphasized predictive factors and offered limited insights into sociocultural processes such as dialogic learning, cognitive development, and learner agency. Similarly, Ajlouni et al. (2023) reported that Jordanian learners appreciated ChatGPT for its scaffolding and feedback. However, concerns about text authenticity and anxiety during tool unavailability pointed to a growing dependence that may hinder learner confidence and agency.

Critically, these studies reflect regional receptiveness to ChatGPT’s pedagogical potential. However, most rely on descriptive data, with limited critical engagement in deeper issues related to learner identity, autonomy, or sustained interaction with the tool (e.g., Alharthi, 2024; Alhammad, 2024; Alotaibi et al., 2025; Haque et al., 2022; Masoudi, 2024). Furthermore, while many studies employ frameworks such as the Technology Acceptance Model (TAM), their application tends to be narrow, and integration with sociocultural perspectives remains largely absent. Notably, the intersection of TAM and sociocultural theory, particularly in understanding how learners engage with AI tools in culturally and linguistically diverse settings, has yet to be meaningfully explored. The present study moves beyond adoption of metrics to investigate how ChatGPT influences learner voice, engagement, and autonomy in EFL Saudi higher education. The following section sheds light on our proposed framework and how it can further enhance our perspective of language learning approaches and the TAM model.

Theoretical Framework

While the field of AI integration in education is rapidly evolving, existing research remains unexplored, especially in its limited incorporation of theoretical frameworks, underrepresentation of learner agency, and insufficient attention to sociocultural and contextual dynamics (Casal & Kessler, 2023; Zheng & Warschauer, 2016). To frame the investigation, this study adopts an integrated theoretical model that combines the Technology Acceptance Model (TAM; Davis, 1989), sociocultural theory (Lantolf & Thorne, 2006; Vygotsky, 1978), languaging concept (Swain & Lapkin, 2013), and learner agency (Benson, 2013; Miller, 2012; Van Lier, 2008). While TAM provides a foundational lens to examine perceived usefulness (PU), ease of use (PEU), and attitude (ATT) as predictors of behavioral intention (BI), it often overlooks the broader learning and educational context. Therefore, sociocultural theory complements TAM by viewing ChatGPT as a mediational instrument that enhances language development through interaction within the given and available Zone of Proximal Development (ZPD; Lantolf & Thorne, 2006; Mahn, 1999). In addition, the concept of languaging highlights how learners use ChatGPT to externalize, think out, and negotiate meaning during tasks such as brainstorming, paraphrasing, and revising (Swain & Lapkin, 2013). Moreover, the study emphasizes learner agency, as students demonstrate autonomy in initiating interactions, customizing responses, and adapting outputs to meet their learning goals (Casal & Kessler, 2023; Wang et al., 2025). This integrative framework offers a more nuanced, socially grounded understanding of ChatGPT use in EFL contexts, moving beyond a purely technological view. However, few studies have applied such a holistic lens, particularly within the Saudi EFL higher education context, making this investigation timely and necessary.

Research Questions

This study investigates the following questions:

What are the perceptions of Saudi higher education EFL learners toward the usefulness of ChatGPT in learning?

What are the attitudes of Saudi higher education EFL learners toward the use of ChatGPT technology as an interactive tool in language learning?

Do the Saudi higher education EFL learners intend to use the ChatGPT technology in language learning?

Methodology

Research Design

The study adopted the explanatory sequential research design by Creswell (2014), where quantitative and qualitative data were gathered in stages, processed separately, and then integrated to provide a more comprehensive understanding. Initially, quantitative data were collected and analyzed. Subsequently, these findings were reinforced and enriched through semi-structured qualitative interviews (Creswell & Plano Clark, 2007). The study was conducted at a public Saudi University during the winter academic year 2023.

Participant and Sampling

This study employed a convenience sampling method for the quantitative phase, as the researchers had access to the participants through their university teaching roles. A total of 181 Saudi English-major students (102 males and 79 females), aged between 20 and 22, were included. Their linguistic proficiency ranged from intermediate to high, and they possessed sufficient English and digital literacy to engage meaningfully with AI-based tools such as ChatGPT. To be included in the study, participants had to confirm that they had used ChatGPT for learning-related purposes. Those who answered “yes” continued to the questionnaire; those who selected “no” were automatically directed to exit. This ensured that only users of ChatGPT were part of the dataset.

To enhance contextual depth, a purposeful subsample of four participants (i.e., three males and a female) was selected from the larger survey pool for semi-structured interviews (see Table 1). These participants met the following inclusion criteria: (i) prior use of ChatGPT for educational purposes and (ii) willingness to engage in recorded interviews. While the number of interviewees was small, the qualitative component aimed not at generalization, but at revealing emotional, ethical, and pedagogical dimensions that may not emerge in surveys. It is well established through classroom interactions and coursework that these students had prior exposure to AI tools, including ChatGPT. AI-related topics were frequently discussed in their language learning and educational technology courses, and all had participated in online AI training sessions.

Table 1.

Interviewee Demographics.

Participant	Gender	English proficiency level	Prior ChatGPT experience	Age range
A	Male	Advanced	Moderate (used occasionally for writing support)	21–23
B	Male	Intermediate	Limited (used once or twice)	20–22
C	Male	Upper-intermediate	Extensive (used regularly for study and translation tasks)	22–24
D	Female	Intermediate	Limited (used briefly for classroom activities)	21–23

Note. Participant labels (A–D) represent pseudonyms to ensure confidentiality and anonymity of the interview data.

Instruments and Data Analysis

To collect quantitative data, we adopted a 12-item questionnaire from Ma et al. (2024) featuring a five-point Likert scale (ranging from strongly agree to strongly disagree). This item was chosen for its impartiality, alignment with our research goals, and accessibility for participants. Ethical approval was obtained, and all participants provided informed consent through signed forms. To maintain confidentiality, each participant in the interviews was assigned a pseudonym.

The questionnaire was divided into three sections: students’ perceptions (4 items), attitudes (4 items), and intentions (4 items) regarding ChatGPT use in higher education. It was uploaded to Google Forms and distributed to all 250 s-year EFL learners at the university. Participants were given 2 weeks to respond. Of the 250 invited, 181 students submitted completed questionnaires, yielding a response rate of 72.4%. Incomplete, inconsistent, or irrelevant responses were excluded from the final analysis. An exclusion question asking whether students had any awareness or use of ChatGPT (“yes” or “no”) ensured that only respondents with experience or knowledge of ChatGPT were included. To ensure construct validity, the items in the survey were reviewed by two experts in educational technology and applied linguistics (Creswell & Creswell, 2018).

To assess reliability, Cronbach’s Alpha was calculated for the three scales, revealing high internal consistency with alpha values above .85, indicating that the items within each scale consistently measured the same concept. A further test was then conducted to confirm the dependability of the scales (see Table 2; Tavakol & Dennick, 2011). Pearson correlation analysis was then performed to examine the relationships between perception, attitude, and intention. Subsequently, multiple regression analysis was conducted to test how well perception and attitude could predict intention. The reliability of each section was assessed using Cronbach’s alpha, yielding scores ranging from .76 to .81 for the perception, attitude, and intention scales, indicating strong internal consistency. Data were saved and analyzed using SPSS version 24.

Table 2.

Reliability Statistics.

Scale	Cronbach’s alpha	No. of items
Perception	$. 876$	4
Attitude	$. 854$	4
Intention	$. 883$	4

Table 1 shows the internal consistency of the survey scales using Cronbach’s Alpha. The results indicate that all three scales used in the study, which are Perception $(α = . 876)$ , Attitude $(α = . 854)$ , and Intention $(α = . 883)$ have high reliability, as values are greater than .70, which are generally considered acceptable. These results suggest that the items within each scale consistently measure the same underlying concept, making the data dependable for further analysis. The regression model was statistically significant, explaining 64.9% of the variance in intention. Perception had a more substantial influence on intention than attitude, as shown by its higher beta value. Scatter plots were also used to confirm the linear relationships between the variables visually. All standard statistical assumptions were checked before analysis to ensure the results were valid and meaningful.

Furthermore, the qualitative data were collected through face-to-face interviews, conducted bilingually in both English and Arabic, over 3 days. These interviews were audio-recorded, transcribed, and analyzed for emerging themes using a directed content analysis approach (Creswell & Plano Clark, 2007), guided by the Technology Acceptance Model (TAM) and the proposed framework. Although only four participants (see Table 1) were interviewed, this number was sufficient to provide meaningful thematic representation that addressed the research questions. Accessing students at the time of data collection was somewhat challenging due to their heavy course loads and limited availability. During the data analysis stage, we initially reviewed each case individually, then revisited and re-examined the key codes highlighted by participants in their narratives. This iterative process allowed for a deeper understanding of emerging patterns through both inductive and deductive reasoning (Terry et al., 2017). We conducted two rounds of coding to identify recurring themes, patterns, distinctions, and similarities in students’ experiences, with a particular focus on their perceptions, attitudes, and intended uses of ChatGPT in educational contexts. Given the bilingual nature of the data, all Arabic excerpts were translated into English and then back-translated by a bilingual expert to ensure linguistic and semantic accuracy and to minimize potential loss of meaning or researcher bias. Two researchers independently coded the data, and any discrepancies were discussed and resolved through consensus to enhance coding reliability and validity. While the small qualitative sample size limits the generalizability of the findings, it provided rich, in-depth insights into students’ lived experiences and nuanced attitudes toward ChatGPT in language learning. To further ensure the validity and trustworthiness of the analysis, a critical friend review of the coding process was conducted.

Research Findings

Numerical Representation of Students’ Perceptions, Attitudes, and Intentions Regarding AI Tools

Perceived Usefulness

The data collected from the first section of the questionnaire revealed that the Saudi EFL learners perceive ChatGPT as highly useful for EFL learning, with an overall mean (M = 3.9), and the standard deviation (SD = .99) depicted that the responses are tightly constructed around the mean. A mean score for the first item (3.9) and (SD = .97) from the respondents’ perceptions of the tool in improving English skills indicates that they are sure that ChatGPT effectively enhances their language proficiency across skills. The mean and SD are repeated for the second item, which categorically names comprehension and communication as the skills improved with the tool, thus reinforcing the earlier result of positive perception of usefulness. ChatGPT also fulfils respondents’ English learning needs (M = 3.9), (SD = 1.02) as reflected in responses to item 3, and is highly perceived as helpful in practicing English (M = 4.0) and (SD = .97). Thus, on all four items that evaluated perceptions of usefulness of ChatGPT technology, the mean scores are in the high range, indicating positive perceptions for this factor. Moreover, the low standard deviations (ranging from .97 to 1.02) proved the high level of agreement toward the usefulness of ChatGPT. The median value of 4 further confirms the consensus among respondents that ChatGPT is a beneficial resource in their learning process.

Table 3 presents the Saudi EFL learners’ perception of ChatGPT’s usefulness in their learning process. The high mean scores (around 4) for all items reflect the higher response trend toward a strong positive perception of ChatGPT’s impact on their English language proficiency. Specifically, the median of 4.0 indicates the students’ stronger approval of the highly conclusive perception. Also, the low standard deviations (around .97–1.02) indicate that the responses have a low variability, which suggests that students agree on the tool’s effectiveness. This ensures the strength of reporting positive perceptions.

Table 3.

Saudi Higher Education EFL Learners’ Perceptions Toward the Usefulness of ChatGPT.

Statement	Mean	Median	SD
1.1. ChatGPT helps me to improve my English language skills	3.9	4	.994444444
1.2. ChatGPT enhances my ability to understand and communicate in English	3.9	4	.976385504
1.3. ChatGPT is valuable for my English language learning needs	3.9	4	1.027238091
1.4. I find ChatGPT to be a valuable tool for practicing English	4	4	.978571421
Average	3.9	4	.994159865

Students’ Attitudes Toward ChatGPT

The survey items presented in Table 3 evaluated respondents’ attitudes, satisfaction, and enjoyment in the use of ChatGPT in language learning. Saudi EFL learners showed high attitudes toward using ChatGPT in learning, with an average mean score of 3.95 and a standard deviation of 0.94. The results for all four items indicate high mean values, and the SD reveals a strong relationship of the agreement. The values show that the respondents find using the technology in language learning convenient and enjoyable. Likely, the ease of use (M = 4.0, SD = .95) contributes to making the learning experience enjoyable; similarly, the mean score and SD for positive attitudes to technology use are also high at 4.0 and .94, respectively, supporting the earlier finding for positive attitudes. Therefore, the learners perceive ChatGPT as helpful and valuable for learning the language.

Table 4 showcases the students’ attitude towards using ChatGPT in language learning. The mean and median values around four illustrate the consistent positive attitude of participants. The standard deviation below 1 represents low variability among students’ responses, indicating that most responses are close. Thus, the student’s attitude toward ChatGPT is positive in their language learning process.

Table 4.

Saudi Higher Education EFL Learners’ Attitudes Towards the Use of ChatGPT.

Statement	Mean	Median	SD
2.1. I am satisfied with using ChatGPT for language learning	4	4	.954634743
2.2. I enjoy using ChatGPT as a language learning tool	3.9	4	.966045442
2.3. I find ChatGPT easy to use and navigate	4	4	.941239312
2.4. Using ChatGPT for language learning is a pleasant experience	3.9	4	.907053779
Average	3.95	4	.942243319

Students’ Intentions to Use ChatGPT

Data in Table 5 shows that Saudi EFL learners intend to use ChatGPT in the future. The participants report an average mean score of 3.7 and an SD of 1.14 regarding their intention to use ChatGPT in their learning. This suggests that most students have a positive intention to continue using ChatGPT. The highest mean score, though marginally higher, was 3.9 and SD 1.11 for their intention to use ChatGPT in future learning situations. This reflects their overall satisfaction with the tool and desire to use it. However, the preference for ChatGPT over other tools was slightly lower, with a mean score of 3.6, and there is variability in SD = 1.20, indicating some students are neutral about it.

Table 5.

Saudi Higher Education EFL Learners’ Intention to Use ChatGPT.

Statement	Mean	Median	SD
3.1. I intend to use ChatGPT for my English language learning in the future	3.99	4	1.120
3.2. I plan to use ChatGPT regularly as a part of my language learning routine	3.89	4	1.111
3.3. I would prefer ChatGPT over other language learning tools if given the choice	3.71	4	1.209
3.4. ChatGPT is an essential tool for my English language learning	3.75	4	1.176
Average	3.83	4	1.142

Table 5 represents the Saudi higher education EFL learners’ intention of using ChatGPT in language learning, compared to perception, and the attitude, where the data variation is noted. The mean value is steadily below 4, reflecting the students’ relatively significant intentions to use ChatGPT in language learning. The standard deviation above 1 reveals the variability among the students’ responses, which suggests that the data sets are a bit spread out from the mean value. It is a minor spread out, and students’ intentions are high toward using ChatGPT in the language learning process.

Relationship Between Perception, Attitude, and Intention

Table 6 shows the Pearson correlation coefficients between Perception, Attitude, and Intention. The analysis revealed strong and statistically significant positive relationships between Perception and Intention (r = .797, p < .001) and between Attitude and Intention (r = .733, p < .001). These findings indicate that as perceptions or attitudes become more positive, participants are more likely to express stronger intentions to use AI tools.

Table 6.

Correlation Coefficient.

Variable Relationship	r	p	N
Perception * intention	.797**	$< . 001$	190
Attitude * intention	$. 733$ **	$< . 001$	189

Correlation is significant at the .01 level (2-tailed).

The result in Table 5 presents the Parson correlation coefficients between Perception, Attitude, and Intention. The results show that there was a strong and statistically significant positive relationship between Perception and Intention $(r = . 797, p < . 001),$ and between Attitude and Intention $(r = . 733, p < . 001)$ . As perception or attitude increases, so does the intention, meaning students with more positive perceptions or attitudes are more likely to show higher intentions. These findings suggest that students with more positive perceptions or attitudes toward ChatGPT are more likely to intend to use it in language learning.

The scatter plots can also visualize this relationship. Figure 1, which displays Intention versus Perception, shows an upward trend of dots, which indicates a strong linear relationship: as perception scores increase from 1 to 5, intention scores also tend to increase—the clustering of points along an upward diagonal shows a strong positive correlation. Similarly, Figure 2, which shows Intention versus Attitude, also presents a clear positive trend, although the spread of the dots is slightly wider compared to Figure 1. This shows a somewhat lower correlation coefficient $(r = . 733)$ , but still strong. Both scatter plots confirm that higher levels of perception and attitude are associated with greater intention. Building on these results, the multiple regression model (Table 6) was statistically significant, explaining 64.9% of the variance in Intention. Perception demonstrated a more substantial influence on intention than attitude among the predictors, as indicated by its higher standardized beta value.

Figure 1.

Scatter plot of intention versus perception.

Figure 2.

Scatter plot of intention versus attitude.

Table 7 shows a multiple regression analysis in which we examined how well Perception and Attitude predict Intention. The overall regression model is statistically significant, $F (2, 186) = 171.581, p < . 001$ , which means that the perception and attitude significantly predict intention. The R² value is .649, meaning that 64.9% of the variance in intention is explained by perception and attitude together, which shows that the regression model is strong. This means the model can effectively predict learners’ intentions in real learning contexts, as perception and attitude together indicate most of the variation in learners’ willingness to integrate ChatGPT into future learning tasks.

Table 7.

Regression Analysis.

Regression model predicting intention	B	SE	β	t	p
Constant	$- . 062$	$. 224$		$- . 278$	$. 781$
Perception	$. 712$	$. 093$	$. 619$	$7.665$	$< . 001$
Attitude	$. 261$	$. 1$	$. 212$	$2.621$	$. 009$

Note. $F (2, 186) = 171.581, p < . 001, R^{2} = . 649$ .

Looking at the individual predictors, perception has a significant positive effect on intention $(B = . 712, t = 7.665, p < . 001) .$ This means that for each one-score increase in perception, intention increases by about .712, while holding attitude constant. The large standardized beta $(β = . 619)$ indicates that perception is the stronger predictor of intention in this model. Similarly, Attitude also significantly predicts intention $(B = . 261, t = 2.621, p = . 009)$ , but its effect is negligible compared to perception. This shows that while attitude still contributes to predicting intention, its influence is weaker than perception. The constant term is not significant $(p = . 781),$ meaning the baseline intention score when both perception and attitude are zero is not significantly different from zero, but this is not usually the focus in interpretation. This regression analysis provides strong evidence that perception and attitude significantly and positively influence intention, with perception being the most impactful factor.

In conclusion, Saudi EFL learners demonstrate positive perceptions and attitudes toward ChatGPT, reflecting clear intentions to continue using it for English language learning. The strong correlations between perception, attitude, and intention support the Technology Acceptance Model (TAM), where perceived usefulness is the primary and critical factor influencing use. From a sociocultural learning perspective, ChatGPT is a mediating tool that supports interactive, scaffolded learning, improving language development through social engagement. According to the statistical data above, learners perceive ChatGPT as useful, hold positive attitudes toward it as an interactive tool, and intend to use it in their language learning journey. These findings highlight ChatGPT’s potential as an effective, sustainable, and engaging educational technology in Saudi EFL contexts. This can also be practically applied to support dynamic interactions with AI and classroom discussions, scaffold language practice, and enhance skills such as writing, reading, and vocabulary in a sustainable manner.

Qualitative Findings

Perceptions of AI-Powered ChatGPT’s Usefulness

Participants largely perceived ChatGPT as an AI-powered tool of value to learners, which enables them to develop language skills; it makes available personalized aid promptly, when required, thus playing a subservient role of mediator agency for facilitating learning processes with due consistency as regards socio-cultural theory that emphasizes desired action as commanded through mediation. For example, Participant A emphasized:

ChatGPT is a useful tool for learning English since it is highly interactive… It helps with vocabulary, grammar, pronunciation, and guides me in writing. It elicits my language needs and supports my learning by breaking down complex tasks.

The quote highlights ChatGPT’s function of mediation externally (not from within the learner), which enables the learner to self-educate with engagement; the AI-generated tool catalyzes zones of proximal development (ZPD) through suitable/needed guidance to the learner community by way of providing learners with formidable tasks, which, of course, are related to language.

Moreover, participants described how ChatGPT mediates thinking and language learning by students as a participative function geared towards extending a supportive role and enabling the learners to structure knowledge together (Learner and AI-powered tool as partners) using an interactive interface. Student B noted:

I use ChatGPT extensively to discuss ideas and generate suggestions. It helps me articulate thoughts I might struggle with alone, making the learning process more interactive and reflective.

Several participants reflected on the dynamic nature of their interaction with ChatGPT, showcasing how the tool supports the generation of meanings and creates basic language-related knowledge; these two (meaning construct and language awareness) are the most important factors in language learning. For instance, Student C stated:

When interacting with ChatGPT, I am not just receiving answers; I am negotiating meaning, clarifying doubts, and reformulating ideas, which enrich my understanding.

However, perceptions were not universally positive. Participant D expressed reservations about the depth of scaffolding ChatGPT can provide:

While it offers help, it sometimes gives ready-made answers that do not encourage me to think critically or solve problems myself.

This critique reflects an awareness of the limitations of AI-language tools, particularly regarding their capacity for higher-level cognitive engagement. These narratives collectively illustrate how learners perceive ChatGPT as both a scaffolding device and a languaging resource that supports their autonomous language learning. This bridges learners’ perceptions of TAM, which extols the tool’s usefulness, with the sociocultural communication theory, which highlights significant limitations. The sociocultural context cannot be overlooked when applying TAM, as social values and cultural influences remain limiting factors in using AI-powered tools. Therefore, socially mediated learning emphasizes integrating social, and cultural dimensions in cognitive development and language interaction.

Attitudes Toward ChatGPT and Its Educational Role

Participants expressed generally positive attitudes toward ChatGPT, recognizing it as a powerful tool that makes available temporary support valuable for facilitating smart language learning, which aligns well with their learning goals. Many appreciated its diverse and multi-dimensional academic performance capabilities from helping with structural writing tasks to generating ideas and providing immediate feedback. Student B shared:

I use ChatGPT for formatting, proofreading, and brainstorming ideas. It feels like having a knowledgeable partner always available to support my writing process.

This indicates a favorable attitude towards the technology’s usability and role as an active, smart, and energetic catalytic AI-powered booster. It provides initial but temporary support for language production and promotes reflective learning. Similarly, Student D noted:

ChatGPT has helped me overcome obstacles when searching for information and organizing my thoughts, which makes learning less stressful.

However, the attitude toward ChatGPT was worrisome as the learners were apprehensive about its efficacy; students simultaneously expressed negativity as they were worried that over-dependence on and continued engagement with it could reduce their level of understanding, affecting their critical thinking ability, and could give a setback to the learning process with lesser application of mind and becoming just copy-cats. For example, Participant A reflected:

Although ChatGPT delivers quick outputs, I worry that depending on it might stop us from developing important skills like critical thinking and problem-solving skills that teachers help nurture through personalized guidance.

The narrative reflects an attitude that values experience with personal application of mind for learning language; if human effort to develop critical thinking and solving difficult situations is not used and temporary support of AI, with continuous dependence as sustained use will create, there is no doubt that language learning, without personal/human effort, real learning will not take place. The user’s mental faculties will not improve, and the degree of skill development will remain static. AI’s support must be withdrawn periodically to develop intensive and deep learning skills. Student C elaborated on this, noting:

While AI is useful, it cannot replace the nuanced feedback from a teacher who understands my learning history, strengths, and weaknesses. The teacher’s role in observing peer interactions and offering tailored advice is irreplaceable.

These reflections reveal learners’ level of criticality and knowledge that socio-cultural factors cannot be ignored while learning a language. Users of AI-generated tools emphasize the importance of the societal communication process and receiving evaluation in context, involving mores, practices, way of living amidst people, at a place, in a situation, objectives of interaction, etc., all conducive and unavoidable for language learning through mental faculty and skill development. The attitude thus encompasses a tension between AI’s role in language learning through its support in solving language learning related issues/problems and keeping intact both agency and cognition.

Furthermore, students’ attitudes revealed a mental condition treating ChatGPT as an AI-generated tool for learning. Users (student-learners) adopt it with self-stated caution, pointing to the need to use it responsibly and considering process-strategies involving active interaction and use. This awareness echoes considering learner-agency like no passive negotiator with socio-cultural dimensions for actively selecting technological applications in the language learning process in users’ learning environment. The sociocultural notion of learner agency is the active negotiation of technology use within their learning ecologies. For instance, Participant D stated:

I appreciate AI’s help but believe that if I let it do too much, it might weaken my ability to think independently and solve problems.

Overall, these attitudes underscore a need to strike a balance between (i) AI’s initial support for language learning and (ii) keeping in constant view the limitations and essential (unavoidable) aid of human mediation in learning. Such a balanced attitude is vital for integrating AI meaningfully into language education without compromising critical engagement and learner autonomy.

Intentions to Use ChatGPT

Participants articulated clear and deliberate intentions regarding their future use of ChatGPT, reflecting an intense motivation and a strong desire to establish an interactive and collaborative relationship with the tool. They expressed a commitment to mastering and integrating AI into their language learning process with confidence and consistency. Student C emphasized the importance of moderation:

I plan to use the AI tool sparingly, as over-reliance on it might undermine the core value of our existence—critical thinking and reflection.

Similarly, Student B expressed a thoughtful balance between efficiency and meaningful interaction:

Although ChatGPT can simplify and speed up tasks, it cannot fully replace the benefits of teacher interaction, who understands my unique strengths and weaknesses.

Participant D described an intentional workflow that preserves independent effort while leveraging AI support:

I intend to craft my work independently first, then use the tool as needed before submitting it for teacher feedback.

Participant A also highlighted using ChatGPT as a temporary support until skills and knowledge are developed to the desired/required extent; this dependency helps develop critical thinking as a joint endeavor and enables overcoming work-related problems accepted as challenges. He reflected,

I plan to use it when I need linguistic support, collaborative thinking, or when a task exceeds my current abilities, seeking further clarification and guidance.

The above responses to the queries reflect learners as cautious users who engage with ChatGPT as autonomous agents rather than passive recipients. Students assert their autonomy on their terms, without becoming unquestioning or uncritical followers. While users may temporarily rely on AI for support, they consciously strive to balance what the tool offers and the socio-cultural demands of communicative and interactive processes.

In conclusion, Saudi EFL learners view ChatGPT positively, recognizing it as a supportive, interactive, and context-sensitive tool for language learning. While learners appreciate its usefulness particularly in grammar, vocabulary, and meaning-making, they also express caution, emphasizing that ChatGPT cannot fully replace the human element. Overreliance, they warn, may stunt mental growth and undermine the development of critical thinking and independent learning skills. Their responses reflect an awareness of the need to maintain cognitive agency, viewing AI tools as temporary scaffolds rather than permanent substitutes. Through (TAM) lens, enriched by sociocultural theory, learners portray ChatGPT as a dynamic support system useful for planning sustained language development while ensuring that human judgment, learner autonomy, and cultural relevance remain central. They advocate for a balanced integration of AI with high-level cognitive engagement, resisting passive consumption and negotiating their interaction with the tool to preserve their active role in learning. In doing so, learners move beyond superficial acceptance and recognize ChatGPT as a dependable, temporary aid that complements rather than replaces human faculties.

Discussion

This study examined Saudi EFL students’ perceptions, attitudes, and intentions toward using ChatGPT in language learning. The findings indicate generally positive perceptions, with students viewing ChatGPT as a supportive and efficient tool, though concerns about over-reliance were also noted. The Technology Acceptance Model (TAM; Davis, 1989) helps explain these findings, particularly by highlighting how perceived usefulness and ease of use shape learners' attitudes and behavioral intentions. However, TAM alone cannot fully account for the sociocultural complexities involved in adopting AI tools for education. To address this gap, the study incorporates sociocultural theory (Vygotsky, 1978) and the concept of learner agency (Miller, 2012), recognizing that learners actively negotiate their use of technology through social, cultural, and contextual factors. These frameworks allowed for a more refined understanding of how ChatGPT functions not just as a tool, but as a mediator of interaction, identity, and meaning-making (Swain & Lapkin, 2013).

Additionally, participants widely perceive ChatGPT as an academically versatile tool, reflecting both enthusiasm for technological integration and a critical awareness of its limitations. As Students A and B demonstrated, participants perceive ChatGPT as a highly useful tool for various academic functions, including proofreading, collaborative thinking, rapid information retrieval, and essay writing assistance. Their widespread AI-mediated use highlights their tech-savviness and eagerness to incorporate technology into their education (Oliver, 2008; Rospigliosi, 2022). This perception aligns with previous research showing that digital-age learners are comfortable with technology from an early age (Acosta-Enriquez et al., 2024; Shin & Lee, 2021).

Furthermore, learners generally show no inhibition toward using technology, associating it with improved learning outcomes and considering tech tools as supportive spaces that scaffold their learning and foster learner agency (Agustini, 2023; Pavlik, 2023). Despite the general enthusiasm, learners do not adopt the technology uncritically. Participants expressed caution, perceiving potential drawbacks such as over-reliance on AI-generated content, which might undermine authentic learning and critical thinking (Raman et al., 2023). These concerns mirror findings from Barrot (2023) and Allehyani and Algamdi (2023), who stress the importance of teacher training to mitigate such risks. Over-reliance concerns include intellectual complacency and ethical issues like plagiarism, reflecting wider academic integrity debates in the AI era (Allehyani & Algamdi, 2023; Perkins, 2023). Participants also demonstrate awareness and agency by using ChatGPT strategically, such as drafting independently before AI assistance. This strategic use of ChatGPT reflects Vygotsky’s sociocultural emphasis on mediated learning, where learners exercise control over tools in ways that extend their cognitive development. By choosing when and how to engage AI assistance, students enact a form of learner agency that balances technological support with personal effort.

Furthermore, participants held predominantly positive attitudes toward ChatGPT, viewing it as a helpful and adaptable tool for language learning. Students A, C, and D expressed confidence in using ChatGPT for future language learning, emphasizing its ability to provide instant information and offer emotional support that reduces learning anxiety (Fathema et al., 2015; Liu & Ma, 2024; Scherer et al., 2019). This emotional support highlights ChatGPT’s dual role not only as a cognitive tool but also as one that helps learners stay motivated and emotionally engaged. Similar findings by Alhammad (2024) and Masoudi (2024) show that Saudi students continue using AI tools to support their understanding and boost confidence. Despite the overall positive outlook, some participants express mixed attitudes, questioning ChatGPT’s long-term effectiveness in improving language skills. This ambivalence resonates with Haque et al. (2022) and Bin-Hady et al. (2024), who observe that initial enthusiasm for technological innovations often coexists with concerns about potential impacts over time. Many students emphasized the need to keep human interaction central in the learning process, valuing teacher support, personal feedback, and social connection, echoing a more human-centered approach to education (Adeshola & Adepoju, 2023; Elbanna & Armstrong, 2024; Strzelecki, 2023). This balanced outlook reflects cautious optimism; students appreciate what ChatGPT offers but are aware of its limits and want to remain active, thoughtful learners (Yu et al., 2024).

Moreover, participants expressed clear intentions to use ChatGPT across a range of academic tasks, including writing, editing, and idea generation. This intention is echoed in prior research by Alhammad (2024) and Masoudi (2024), who found Saudi students aim to incorporate AI tools to enhance learning and comprehension sustainably. Their approach reflects active learner agency and critical digital literacy, exemplified by strategies such as independent drafting before AI-assisted revision. Although students exhibit strong intentions to continue using ChatGPT, they remain mindful of its limitations and emphasize the importance of balancing technological support with human educational elements (Rogers, 2003). This careful, reflective stance points to a more pragmatic and responsible approach to using AI in language learning, one that values both innovation and human-centered education.

Comparatively speaking, the study findings reveal similar sentiments to those of Agustini (2023) on a global scale, particularly in cultivating student agency and fostering positive attitudes toward using AI in English language learning. In some ways, the findings also expand on a local level what Alharthi (2024) reported regarding emotions and cultural considerations of AI as a facilitating tool in language learning. While the studies by Alharthi (2024), Ajlouni et al. (2023), and Alhammad (2024) trace the use of AI across Saudi universities and explore students’ perceptions and awareness in Jordan, our findings reflect only one context but with a greater focus on rich, detailed insights. Building on this perspective, Alotaibi et al. (2025) highlight that learners demonstrate keener exercise of agency and can engage in identity construction through languaging by using technology. They employ AI tools as mediating instruments situated socioculturally within the specific educational and cultural context of Saudi Arabia’s higher education system. This approach underscores how AI can be integrated thoughtfully within language learning, respecting both the sociocultural environment and the unique needs of Saudi learners. However, according to the study data, the very features that render ChatGPT “effective” in Saudi higher education, its speed, accessibility, and expansive content, may paradoxically constrain learner agency and meaningful languaging if not situated within pedagogical frameworks that prioritize critical engagement, identity negotiation, and socioculturally responsive practices.

All in all, the tension between AI’s benefits and risks calls for a reevaluation of EFL traditional teaching practices. Educators are encouraged to design learning environments that foster and sustain critical thinking, reflection, and creativity instead of relying on rote memorization or passive content delivery. By exploring students’ agency and languaging skills in real educational contexts, the data can reveal how learners actively negotiate meaning, mediate understanding, and co-construct knowledge while engaging with technology. This can expand TAM to better account for the dynamic, socially situated nature of technology use in learning, integrating cognitive, affective, and sociocultural dimensions. Consequently, such research can inform the design of more effective technology-based learning theories and pedagogies that emphasize learner empowerment, interaction, and critical engagement rather than passive adoption of tools (Kolade et al., 2024).

Conclusion and Implication

Overall, AI technologies like ChatGPT have the potential to enhance educational quality and adapt to future educational and personal needs. From the perspective of the Technology Acceptance Model (Davis, 1989), learners’ perceptions of usefulness and ease of use fundamentally shape their willingness to adopt such tools. With proper training and ethical implementation, AI technologies can unlock new educational opportunities and foster a more inclusive and democratic education system, moving beyond traditional teacher-dominant paradigms (Bin-Hady et al., 2024; Oliver, 2008). Teachers of English can incorporate AI-mediated tools in their daily courses for brainstorming, processing informational or intellectually demanding content, and for fostering students’ agency in language learning. However, sociocultural theory (Vygotsky, 1978) reminds us that effective integration requires understanding AI tools as socially mediated artifacts that learners actively appropriate within their cultural and educational contexts. This study highlights the importance of balancing technological integration with critical awareness and responsible use, ensuring that AI tools such as ChatGPT serve as supplements rather than substitutes for human-driven educational practices. While the readiness of learners and the technology’s ease of use suggest that ChatGPT could be a valuable educational tool, its potential drawbacks must be further explored—particularly in relation to class, gender, and access.

Findings suggest a promising potential for integrating ChatGPT into English language learning in Saudi Arabia, reflecting an emerging trend in instructional and scaffolding approaches as discussed in recent studies on the shift toward the AI era (Raman et al., 2023; Rospigliosi, 2023). At a macro level, institutions and policymakers should advocate for the responsible, ethical, and productive use of AI tools rather than imposing blanket prohibitions. Implementing explicit policies and developing alternative assessment strategies can help manage the disruptive consequences of excessive ChatGPT use. As our findings suggest, policymakers and administrators should recognize ChatGPT’s significant role as a resource, uphold academic integrity, and promote responsible use (Alowedi & Al-Ahdal, 2023).

Recommendations and Limitations

To support the integration of ChatGPT into EFL contexts, institutions should offer hands-on training for students to explore its features and use it effectively for feedback and sustainable learning. Seminars and online forums are also recommended to help instructors share best practices and allow students to discuss experiences and ethical concerns. Course designers should embed AI-supported tasks into curricula, and students should be encouraged to engage with ChatGPT as part of their ongoing learning. This study has some limitations. It relied on an online survey that may not fully represent the broader student population, and the interview sample was limited to four participants, possibly narrowing the range of perspectives. Moreover, as students only responded to survey and interview questions, these perspectives may reflect socially desirable attitudes rather than real behavior. Additionally, since ChatGPT and similar AI tools are continuously evolving, the findings may have limited long-term relevance. Future research should employ more representative and randomized sampling methods, consider longitudinal designs to capture dynamic perceptions and uses of AI tools in English education, and include experimental approaches to further unpack learners actual behavior.

Footnotes

ORCID iD

Majed Abdullah Alharbi

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors extend the appreciation to the Deanship of Postgraduate Studies and Scientific Research at Majmaah University for funding this research work through the project number NO. R-2025-1946.

Declaration of Conflicting Interests

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

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.

References

Abdaljaleel

Barakat

Alsanafi

Salim

N. A.

Abazid

Malaeb

Sallam

(2024). A multinational study on the factors influencing university students’ attitudes and usage of ChatGPT. Scientific Reports, 14(1), Article 1983. https://doi.org/10.1038/s41598-024-52549-8

Acosta-Enriquez

B. G.

Arbulú Ballesteros

M. A.

Huamaní Jordan

López Roca

Saavedra Tirado

(2024). Analysis of college students’ attitudes toward the use of ChatGPT in their academic activities: Effect of intent to use, verification of information and responsible use. BMC Psychology, 12(1), Article 255. https://doi.org/10.1186/s40359-024-01764-z

Adeshola

Adepoju

A. P.

(2023). The opportunities and challenges of ChatGPT in education. Interactive Learning Environments, 32(10), 6159–6172. https://doi.org/10.1080/10494820.2023.2253858

Agustini

N. P. O.

(2023). Examining the role of ChatGPT as a learning tool in promoting students’ English language learning autonomy relevant to Kurikulum Merdeka Belajar. Edukasi-a: Jurnal Pendidikan dan Pembelajaran, 4(2), 921–934. https://doi.org/10.62775/edukasia.v4i2.373

Ajlouni

A. O.

Wahba

F. A.-A.

Almahaireh

A. S.

(2023). Students’ attitudes towards using ChatGPT as a learning tool: The case of the University of Jordan. International Journal of Interactive Mobile Technologies, 17(18), 99–117. https://doi.org/10.3991/ijim.v17i18.41753

Alhammad

A. I.

(2024). The impact of ChatGPT in developing Saudi EFL learners’ literature appreciation. World Journal of English Language, 14(2), Article 331. https://doi.org/10.5430/wjel.v14n2p331

Alharthi

S. M.

(2024). Beyond traditional language learning: EFL student views on ChatGPT in Saudi Arabia. Arab World English Journal, 15(10), 15–35. https://doi.org/10.24093/awej/call10.2

Aljabr

F. S.

Al-Ahdal

A. A. M. H.

(2024). Ethical and pedagogical implications of AI in language education: An empirical study at Ha’il University. Acta Psychologica, 251, Article 104605. https://doi.org/10.1016/j.actpsy.2024.104605

Allehyani

S. H.

Algamdi

M. A.

(2023). Digital competences: Early childhood teachers’ beliefs and perceptions of ChatGPT application in teaching English as a second language. International Journal of Learning, Teaching and Educational Research, 22(11), 343–363. https://doi.org/10.26803/ijlter.22.11.19

10.

Alotaibi

H. M.

Sonbul

S. S.

El-Dakhs

D. A.

(2025). Factors influencing the acceptance and use of ChatGPT among English as a foreign language learners in Saudi Arabia. Humanities and Social Sciences Communications, 12, Article 628. https://doi.org/10.1057/s41599-025-04945-2

11.

Alowedi

N. A.

Al-Ahdal

A. A. M. H.

(2023). Artificial intelligence based Arabic-to-English machine versus human translation of poetry: An analytical study of outcomes. Journal of Namibian Studies: History Politics Culture, 33, 1523–1538. https://doi.org/10.20372/jns.2023.33.10

12.

Barrot

J. S.

(2023). Using ChatGPT for second language writing: Pitfalls and potentials. Assessing Writing, 57, Article 105. https://doi.org/10.1016/j.asw.2023.100745

13.

Benson

(2013). Teaching and researching: Autonomy in language learning (2nd ed.). Routledge.

14.

Benuyenah

(2023). Commentary: ChatGPT use in higher education assessment: Prospects and epistemic threats. Journal of Research in Innovative Teaching & Learning, 16(1), 134–135. https://doi.org/10.1108/JRIT-03-2023-097

15.

Bin-Hady

W. R. A.

Ali

J. K. M.

Al-Humari

M. A.

(2024). The effect of ChatGPT on EFL students’ social and emotional learning. Journal of Research in Innovative Teaching & Learning, 17(2), 243–255. https://doi.org/10.1108/JRIT-02-2024-0036

16.

Creswell

J. W.

(2014). Research design: Qualitative, quantitative, and mixed methods approaches (4th ed.). SAGE Publications.

17.

Creswell

J. W.

Creswell

J. D.

(2018). Research design: Qualitative, quantitative, and mixed methods approaches (5th ed.). SAGE Publications.

18.

Creswell

J. W.

Plano Clark

V. L.

(2007). Designing and conducting mixed methods research. Sage.

19.

Casal

J. E.

Kessler

(2023). Can linguists distinguish between ChatGPT/AI and human writing? A study of research ethics and academic publishing. Research Methods in Applied Linguistics, 2(3), Article 68. https://doi.org/10.1016/j.rmal.2023.100068

20.

Davis

F. D.

(1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319–339. https://doi.org/10.2307/249008

21.

Elbanna

Armstrong

(2024). Exploring the integration of ChatGPT in education: Adapting for the future. Management & Sustainability: An Arab Review, 3(1), 16–29. https://doi.org/10.1108/MSAR-03-2023-0016

22.

Fathema

Shannon

Ross

(2015). Expanding the technology acceptance model (TAM) to examine faculty use of learning management systems (LMSs) in higher education. Journal of Online Learning and Teaching, 11(2), 210–232.

23.

Haque

M. U.

Dharmadasa

Sworna

Z. T.

Rajapakse

R. N.

Ahmad

(2022). “I think this is the most disruptive technology” Exploring sentiments of ChatGPT early adopters using Twitter data. Computer Science: Computation and Language, 17, 320–332. https://doi.org/10.48550/arXiv.2212.05856

24.

Holmes

Anastopoulou

Schaumburg

Mavrikis

(2018). Technology-enhanced personalised learning: Untangling the evidence. Robert Bosch Stiftung.

25.

Hutson

Jeevanjee

Vander Graaf

Lively

Weber

Weir

(2022). Artificial intelligence and the disruption of higher education: Strategies for integration across disciplines. Creative Education, 13(12), 3953–3980. https://doi.org/10.4236/ce.2022.1312253

26.

Kolade

Smith

Johnson

(2024). Artificial intelligence in lifelong learning: Opportunities for AI-assisted assessment. International Journal of Educational Technology in Higher Education, 21(1), 1–17. https://doi.org/10.1186/s41239-024-00331-7

27.

Lantolf

J. P.

Thorne

S. L.

(2006). Sociocultural theory and the genesis of second language development. Oxford University Press.

28.

Liu

(2024). Measuring EFL learners’ use of ChatGPT in informal digital learning of English based on the technology acceptance model. Innovation in Language Learning and Teaching, 18(2), 125–138. https://doi.org/10.1080/17501229.2023.2240316

29.

Wang

Pang

X. S.

Wang

(2024). Exploring user adoption of ChatGPT: A technology acceptance model perspective. International Journal of Human–Computer Interaction, 29(5), 1–15. https://doi.org/10.1080/10447318.2024.2314358

30.

Mahn

(1999). Vygotsky’s theory of social development and constructivism. In Lee

C. D.

Smagorinsky

(Eds.), Vygotskian perspectives on literacy research: Constructing meaning through collaborative inquiry (pp. 39–64). Cambridge University Press.

31.

Malik

Khan

Hussain

(2023). How is ChatGPT transforming academia? Examining its impact on teaching, research, assessment, and learning. SSRN Electronic Journal, 10(1), 15–30. https://doi.org/10.2139/ssrn.4413516

32.

Masoudi

H. A.

(2024). Effectiveness of ChatGPT in improving English writing proficiency among non-native English speakers. International Journal of Educational Sciences and Arts, 3(4), 62–84. https://doi.org/10.59992/IJESA.2024.v3n4p2

33.

Miller

E. R.

(2012). Agency in the making: Adult immigrants’ accounts of language learning and work. TESOL Quarterly, 46(3), 407–432.

34.

Moldt

J.-A.

Festl-Wietek

Madany Mamlouk

Nieselt

Fuhl

Herrmann-Werner

(2023). Chatbots for future docs: Exploring medical students’ attitudes and knowledge towards artificial intelligence and medical chatbots. Medical Education Online, 28(1), Article 219. https://doi.org/10.1080/10872981.2023.2182659

35.

Oliver

(2008). Adult education as a means of democratization. REPORT-Zeitschrift für Weiterbildungsforschung, 31(2), 18–27.

36.

Pavlik

J. V.

(2023). Collaborating with ChatGPT: Considering the implications of generative AI. Journalism & Mass Communication Educator, 87(1), Article 107769582211495. https://doi.org/10.1177/10776958221149577

37.

Perkins

(2023). Academic integrity considerations of AI large language models in the post-pandemic era: ChatGPT and beyond. Journal of University Teaching & Learning Practice, 20(2), 1–24. https://doi.org/10.53761/1.20.02.07

38.

Raman

Mandal

Das

Kaur

Sanjanasri

J. P.

Nedungadi

(2023). University students as early adopters of ChatGPT: Innovation diffusion study. Journal of Educational Technology, 15(4), 235–250. https://doi.org/10.21203/rs.3.rs-2734142/v1

39.

Rogers

E. M.

(2003). Diffusion of innovations. Free Press.

40.

Rospigliosi

P. A.

(2022). Why we need critical pedagogy in post-pandemic interactive learning environments. Interactive Learning Environments, 30(5), 779–781. https://doi.org/10.1080/10494820.2021.1964472

41.

Rospigliosi

P. A.

(2023). Artificial intelligence in teaching and learning: What questions should we ask of ChatGPT? Interactive Learning Environments, 31(1), 1–3. https://doi.org/10.1080/10494820.2022.2147754

42.

Scherer

Siddiq

Tondeur

(2019). The technology acceptance model (TAM): A meta-analytic structural equation modeling approach to explaining teachers’ adoption of digital technology in education. Computers & Education, 128(3), 13–35. https://doi.org/10.1016/j.compedu.2018.09.009

43.

Shin

J. H.

Lee

J. W.

(2021). Athlete brand image influence on the behavioral intentions of technology in education. Computers & Education, 120, 13–35. https://doi.org/10.1016/j.compedu.2018.12.004

44.

Strzelecki

(2023). To use or not to use ChatGPT in higher education? A study of students’ acceptance and use of technology. Interactive Learning Environments, 32(9), 5142–5155. https://doi.org/10.1080/10494820.2023.2238237

45.

Sviridova

Yastrebova

Bakirova

Rebrina

(2023). Immersive technologies as an innovative tool to increase academic success and motivation in higher education. Frontiers in Education, 8, Article 11. https://doi.org/10.3389/feduc.2023.1192760

46.

Swain

Lapkin

(2013). A Vygotskian sociocultural perspective on immersion education: The L1/L2 debate. Journal of Immersion and Content-Based Language Education, 1(1), 101–129. https://doi.org/10.1075/jicb.1.1.05swa

47.

Taecharungroj

(2023). “What can ChatGPT do?” Analyzing early reactions to the innovative AI chatbot on Twitter. Big Data and Cognitive Computing, 7(1), Article 35. https://doi.org/10.3390/bdcc7010035

48.

Tavakol

Dennick

(2011). Making sense of Cronbach’s alpha. International Journal of Medical Education, 2, 53–55. https://doi.org/10.5116/ijme.4dfb.8dfd

49.

Taylor

D. L.

Yeung

Bashet

A. Z.

(2021). Personalized and adaptive learning. In Ryoo

Winkelmann

(Eds.), Innovative learning environments in STEM higher education: Opportunities, challenges, and looking forward (pp. 17–34). Routledge.

50.

Terry

Hayfield

Clarke

Braun

(2017). Thematic analysis. In Willig

Rogers

W. S.

(Eds.), The SAGE handbook of qualitative research in psychology (pp. 17–36). SAGE Publications Ltd.

51.

Tuomi

(2019). The impact of artificial intelligence on learning, teaching, and education: Policies for the future (JRC Science for Policy Report). European Commission.

52.

Vygotsky

L. S.

(1978). Mind in society: The development of higher psychological processes. Harvard University Press.

53.

Van Lier

(2008). Agency in the classroom. In Lantolf

J. P.

Poehner

M. E.

(Eds.), Sociocultural theory and the teaching of second languages (pp. 163–186). Equinox.

54.

Wohlfart

Wagner

(2023, August 22–25). Digital transformation? A longitudinal interview study on teachers’ acceptance and usage of digital tools in times of COVID-19 [Conference session]. European Conference on Educational Research, Glasgow, UK.

55.

Wang

Yuan

Moon

Kim

K. H.

(2025). Enhancing parasocial brand experience and brand equity: Unleashing the advantages of ChatGPT service. International Journal of Consumer Studies, 49(2), Article e70029. https://doi.org/10.1111/ijcs.70029

56.

S.-C.

Huang

Y.-M.

T.-T.

(2024). Tool, threat, tutor, talk, and trend: College students’ attitudes toward ChatGPT. Behavioral Sciences, 14(9), Article 755. https://doi.org/10.3390/bs14090755

57.

Zheng

Warschauer

(2016). Learning in one-to-one laptop environments: A meta-analysis and research synthesis. Review of Educational Research, 86(4), 1052–1084. https://doi.org/10.3102/0034654316628645