1. Introduction

Research has recently focused on integrating technologies into language teaching and assessment in the context of English language teaching (Rusmiyanto et al., 2023). Among the various technological innovations, the emergence of Artificial Intelligence (AI) tools has offered novel opportunities and challenges for educators worldwide. One such tool, ChatGPT, has become a focal point of pedagogical innovation for its potential applications in language instruction and assessment. ChatGPT, as an AI-powered tool, can produce human-quality content, answer questions in an informative manner, and translate different languages (Sufyan et al., 2024). This raises fascinating possibilities for incorporating AI into language learning settings, unlocking the numerous opportunities to cultivate core linguistic competencies, including speaking, listening, reading, and writing. Moreover, innovations in AI-driven technologies, including speech recognition systems, conversational chatbots, virtual tutoring platforms, and specialized language applications, have surfaced as transformative tools. They offer learners dynamic and engaging experiences. Features such as instantaneous feedback, customized assessments, and tailored learning paths are instrumental in augmenting communicative skills and expediting the mastery of new languages (Rusmiyanto et al., 2023). Nevertheless, more studies are still required to explore the potential advantages and disadvantages of ChatGPT applications in such settings. Although the proponents support the potential advantages of these virtual technologies, others are concerned about the challenges and possible adverse effects of technology-based approaches (Fernández-Batanero et al., 2021). Besides, although researchers show an increasing interest in the ChatGPT applications in a language teaching context, scant research has examined the perceptions of English language teachers, especially in cross-cultural settings. In addition, one of the critical issues is dealing with the perceptions of teachers directly experiencing the strengths and drawbacks of the ChatGPT applications in language teaching strategies. More precisely, research should seek to highlight Iranian and non-Iranian English teachers’ perceptions of ChatGPT usage in language instruction and assessment and its likely contribution to updated educational methodologies and dealing with possible challenges related to the implementation of this technology. 

Despite the potential effect of AI in the field of education, further research is essential to examine the EFL teachers’ perceptions toward the effectiveness and challenges of AI-powered tools such as ChatGPT. Meanwhile, the successful integration of AI in education depends on several factors, one of which is the teachers’ perspective. As teachers around the world try to keep pace with modern advancements, the implementation of AI-driven tools such as ChatGPT into language instruction and assessment has become a key focus. Equally essential is obtaining an understanding of the perspectives of Iranian teachers regarding the usefulness and challenges of ChatGPT in EFL classes. In the realm of education, understanding and comparing the perceptions of both Iranian and non-Iranian teachers on AI integration is of significant importance given their vital role in conducting successful instruction, appropriate assessment, and facilitating the process of language learning.

The 21st-century technological advancements have significantly affected various sectors such as language instruction and assessment. Many educators discuss the emergence of ChatGPT in educational contexts and the potential changes it might bring to teaching strategies. The proponents of such technologies as ChatGPT and similar AI tools emphasize their contribution to teaching and learning. Baskara (2023) believes that learners can take advantage of ChatGPT to develop language skills like writing and vocabulary acquisition while being exposed to personalized practice materials and explanations. Moreover, Hong (2023) asserts that ChatGPT can bring interactive learning experiences to learners, making language practices more engaging and fun. ChatGPT is available, allowing students to practice English conversation anytime and anywhere, making learning more convenient and accessible, as well as providing immediate feedback on grammar, vocabulary usage, and sentence structure, helping students learn from their mistakes in real-time (Hawanti & Zubaydulloevna, 2023). Overall, based on the majority of studies, the impact of chatbots such as ChatGPT on English language education can be considered positive (Esfandiari & Ghamari, 2025).

In the domain of English as a Foreign Language (EFL) instruction, while ChatGPT suggests a plethora of advantages, it is imperative to acknowledge the inherent challenges and limitations that accompany its application. Tan and Subramonyam (2024) highlight the potential for ChatGPT to deliver responses that lack cultural relevance or contextual appropriateness, posing a risk of misunderstandings, particularly where the subtleties of culture play a crucial role in language learning. Furthermore, Sharma and Yadav (2022) caution against an overreliance on such technology, which may impede genuine communicative experiences with native speakers that are vital for linguistic proficiency. Additionally, the susceptibility of ChatGPT to errors, as noted by Hua et al. (2024), could inadvertently solidify linguistic inaccuracies if not diligently supervised by educators. Concerns surrounding privacy and data security also emerge, especially when sensitive information is processed or retained by the system, as indicated by Huallpa (2023). Lastly, the absence of emotional intelligence in ChatGPT, as Wang et al. (2023) observe, means it cannot replicate the emotional support and motivation typically provided by human instructors, an element deemed essential in language learning environments. It is, therefore, crucial for educators to employ ChatGPT judiciously and in tandem with established pedagogical practices to counterbalance these drawbacks and foster a comprehensive and efficacious approach to EFL instruction. In general, the challenges are mainly classified as concerns about accuracy and the presence of misinformation, the risk of plagiarism and its effect on critical thinking, privacy and ethical concerns, cultural and social concerns, pedagogical and technical challenges, concerns regarding replacing human interaction in language acquisition, and the requisite for continuous improvement, updates, and monitoring (Esfandiari & Ghamari, 2025).

Recent years have witnessed significant transformations in EFL instruction, remarkably affected by the development of AI technologies, including ChatGPT. Hence, both instructors and learners face novel possibilities in the context of personalized language learning, real-time language practice, and lesson planning. The task of analyzing students’ data and subsequently adapting lesson plans and content is carried out by AI-powered platforms in personalized language learning. For instance, learners' strengths and weaknesses are discovered by AI algorithms in platforms such as Duolingo, based on which targeted exercises are provided to ensure meeting the learners’ educational needs (Zargoun, 2024) and guide each student’s learning process at a personalized pace.

Recent reports have highlighted the use of ChatGPT in language learning, as it provides personalized guidance to learners and enables educators to create instructional materials tailored to the individual needs and interests of their students. This personalized approach can enhance learner engagement and improve the effectiveness of language learning. Moreover, ChatGPT can design practical tasks that align with students’ proficiency levels and learning objectives, thereby enhancing language skill acquisition and increasing learners’ motivation (Baskara & Mukarto, 2023). Another significant feature is ChatGPT’s ability to generate authentic language content, such as dialogues, articles, and reading texts, which can improve learners’ reading comprehension and provide a more authentic learning context. Chatbots foster a deeper understanding of the language, facilitating instantaneous interaction and motivating learners to refer to their linguistic capabilities in authentic contexts, which will enhance their command of the language while also improving their self-confidence in practically employing the language in real-world conditions. According to Amin (2023), AI's capability of rendering prompt evaluations on various dimensions, including pronunciation, grammar, and lexicon, underlies the immediacy of language practice. Kiryakova and Angelova (2023) in a study concluded that university professors from Trakia University in Bulgaria have a general positive attitude toward the implementation of ChatGPT in their instruction (41.4%). They consider ChatGPT as a tool to support time-consuming teaching activities (60.9%), improve interest, activate and engage learners (59.8%), and stimulate their critical thinking and creativity (47.1%).  However, some faculty members remain unfamiliar with ChatGPT and are therefore uncertain about whether it represents a threat or a valuable opportunity.

One of the main challenges many EFL educators face is how to design lesson plans that are engaging and effective on the one hand and compatible with educational goals on the other hand, both of which can come true to a significant extent using AI and ChatGPT technologies. The ability of such tools to generate customized lesson plans, materials, and exercises to target requirements and interests of the learners makes them of special importance. For example, teachers can ensure the provision of challenging content using AI-generated lesson plans, which are adjustable to learners’ progress without overwhelming them. In addition, the learning experience can be enhanced using AI, as relevant reading materials, multimedia resources, and real-world examples are provided to improve learners’ mastery of language (Onal & Kulavuz-Onal, 2024). Thus, educators can save time using these technologies while ensuring that relevant and up-to-date lesson plans are compatible with learners’ needs. Several valuable possibilities of AI and ChatGPT, including personalized language learning, real-time language practice, and indispensable assistance in lesson planning, have reshaped the landscape of EFL language teaching, enabling educators to generate customized learning experiences, facilitating real-world language applications, and optimizing lesson planning, eventually improving the language instruction quality. All these advancements and technological innovations have enabled EFL educators and learners to take advantage of the power of AI and ChatGPT to achieve the highest quality of language learning (Amin, 2023). Yang’s comparative study (2025) revealed significant quantitative and qualitative differences between AI, human, and no teaching assistant conditions in a language teaching setting. Quantitatively, AI-assisted learning led to the most stable gains, with students achieving a mean score of 24.23, outperforming those with human assistants (M = 20.26) and those with no assistance (M = 16.30). Qualitatively, AI was admired for its availability and personalized feedback, but lacked emotional connection. Human assistants provided emotional support and contextualized explanations, but had availability challenges.

AI plays a significant role in EFL assessment by providing automated tools for evaluating language proficiency, providing personalized feedback, and tracking student progress. AI systems can analyze written and spoken language, detect errors, and provide suggestions for improvement. These tools can help teachers save time on grading and provide more targeted support to students (Schmidt-Fajlik, 2024).

Assessment and evaluation have undergone considerable transformations with the emergence of AI technologies and their integration into education (Amin, 2023). According to Moqbel and Al-Kadi (2023), the educational assessment landscape has been currently reshaped by the capabilities of ongoing formative assessment, AI-enhanced performance analysis, and the evolving contribution of teachers to AI-generated work evaluation in the AI era. Amin (2023) emphasizes that such systems are capable of identifying patterns and trends, enabling instructors to look at individual and class-wide performance from a new standpoint. One example is when educators modify their teaching strategies and offer timely interventions, as AI detects students are struggling with particular topics or skills. Besides, a more holistic view of student progress is provided by AI-enhanced performance analytical procedures, which go beyond traditional grade-centered assessments, enhancing teachers’ role and providing valuable insights and real-time feedback. As a result, the prevalence of AI-generated assessments and evaluations will help instructors play an essential role in directing the learning process and clarifying the outcomes. While ChatGPT is a powerful tool, it should not replace the role of the instructor in assessment. Educators must improve their assessment abilities to make sure that learners receive corrective feedback that is not only accurate but also contextually relevant and personalized (Baleghizadeh & Saeedi, 2024).

Previous studies identified the benefit of language learning through AI-powered chatbots, demonstrating how they help to improve pronunciation (Liakin et al., 2014), grammar (Al-Kaisi et al., 2019), and vocabulary (Chen et al., 2020). The pace at which AI is developing means that English language teachers will have to redefine their roles and modify their pedagogy to integrate new technologies into their practice, ultimately achieving better learning results (Guo & Wang, 2024).

Practical insights can also be provided by this comparative study into the effects of cultural and pedagogical backgrounds on the AI acceptance and application in language teaching and assessment. In the meantime, the research results can have significant implications for curriculum development, teacher training, and policy-making to ensure that the employment of AI tools, including ChatGPT, takes into account learners’ needs and teachers’ expectations in different settings. Hence, the study seeks to compare Iranian and non-Iranian English teachers’ views toward the application of ChatGPT in language teaching and assessment through an investigation of their perceptions. In this regard, the following questions are raised and sought to be answered.

·         Q1. Are there any statistically significant differences between Iranian and non-Iranian EFL teachers’ perceptions of ChatGPT-associated knowledge and experiences?

·         Q2. Are there any statistically significant differences between Iranian and non-Iranian EFL teachers’ perceptions of the effectiveness and usefulness of ChatGPT in language instruction and assessment?

·         Q3. Are there any statistically significant differences between Iranian and non-Iranian EFL teachers’ perceptions regarding the challenges of employing ChatGPT in language instruction and assessment?

2. Literature Review

AI has touched every part of human life, with applications as diverse as intelligent devices, production, transportation, medicine, and many more. With the technology’s development speeding ahead, applications expand continuously without any discernible limits. One field where AI has contributed immensely is education, particularly English language teaching and assessment (Bin & Mandal, 2019). The development of AI technologies like ChatGPT is outpacing research into their learning effects, especially in teaching and assessing the English language. While studies into AI applications in most domains, including learning, have been carried out, there is limited research into the application of generative AI models like ChatGPT in the case of EFL teachers’ perceptions toward English language teaching and assessment. So, there is a need to fill the gap and critically examine the effectiveness and challenges encountered by ChatGPT. More research is required to map the present research context, determine the topics to investigate in more detail, and maximize its potential to contribute to English language instruction (Locky Law, 2024).

The current research aimed to meet these requirements by investigating the perspectives of Iranian and non-Iranian EFL teachers toward the challenges and possibilities presented by ChatGPT, offering insights to facilitate more effective utilization of this technology in English language teaching and assessment. Kamnis (2023) defines a Generative Pre-trained Transformer (GPT), released by OpenAI (San Francisco, California) in 2018, as a type of Large Language Model (LLM) seeking to replicate human language processing abilities. Through GPT, deep learning is leveraged, and powerful algorithms are provided to carry out different language-associated tasks, including text generation, question answering, and translation, while offering context comprehension capabilities to generate human language-like responses (Baidoo-Anu & Owusu Ansah, 2023). 

ChatGPT-4 language models were launched by OpenAI in March 2023 (Skavronskaya et al., 2023), which respond based on their perception of the language and context of the prompts, contributing to coherent and contextually related conversations (Fu et al., 2023). Given that, researchers are increasingly paying attention to large language models in education, examining their broad range of applications, particularly considering the availability of other similar options, including Bing Chat, Bard, and Ernie (Rudolph et al., 2023). A variety of applications have been reported for large language models, providing the required assistance as perceived by the learners throughout the learning process.

Hence, the potential advantages and challenges related to their application require careful consideration. While there have been some research studies investigating the applications of AI in various domains, such as education, research studies specifically examining the use of AI models, such as ChatGPT, in English language teaching and assessment remain relatively scarce. An example is the study carried out by Gordani et al. (2023), which employed a randomized controlled method involving 60 English language learners from Salman Farsi University of Kazerun. The study divided participants into experimental and control groups, with the experimental group receiving chatbot-integrated instruction and the control group being taught in the same period using the traditional method. The findings showed that the chatbot-based approach significantly enhanced learners' language speaking abilities, motivation, and overall attitudes toward English language learning when compared to traditional teaching methods. Asgari Vaziri and Asgari Goudarzi (2023) conducted a study exploring the applications of AI in English language teaching, focusing on areas such as grammar error detection and correction, assessment of listening and speaking skills, personalized learning recommendations, and interaction with talking robots. While their findings revealed that AI can enhance student engagement, accelerate learning, and improve language proficiency, they also highlighted concerns related to privacy, technological dependency, and bias. In another study, Mousavi (2023) explored innovative methods of integrating AI into English language teaching to enhance the learning process. The findings indicated that the integration of AI in English language instruction can significantly impact the learning process and help students improve their language skills.

3. Research Method

English teachers' ChatGPT-associated knowledge, perceptions, and experiences were investigated using an online survey through Google Forms, which also examined the participants’ perceptions of its advantages and drawbacks. Convenience sampling was used to select participants based on their accessibility. This led to a sample of 93 English language teachers (45 Iranian and 48 non-Iranian), who anonymously participated in the study after giving informed consent. Once the researchers finished using the Google Survey Creator and completed the questionnaire, they sent the scale via a link and shared it using LinkedIn. Additionally, the researchers sent the survey to Iranian and non-Iranian teachers’ email addresses, and a sample of 93 Iranian and non-Iranian English teachers completed the questionnaire. The participants belonged to various countries like Iran, Malaysia, Turkey, Azerbaijan, and Russia, of whom 76 males and 17 females (81.7% and 18.3%, respectively) completed the survey. Considering the differences among countries, the strategy of “different social systems, different educational outputs” was taken into consideration. The participants of the study were novice and experienced male and female teachers with the age range of 20 to above 50 teaching in language institutes and universities. In terms of their level of education, the majority of both Iranian and non-Iranian teachers held a Master's degree, and a slightly higher percentage of non-Iranian teachers held a PhD (10.4% vs. 4.4%).

The questionnaire had two sections, one collecting demographic data, including gender, age, teaching experience, country, and academic qualifications, and the other investigating participants' perceptions of ChatGPT and integrating this tool into language instruction and assessment methods. The questionnaire items were designed with due care to examine different ChatGPT aspects, such as knowledge of ChatGPT, ChatGPT effectiveness, and ChatGPT-related challenges. The questionnaire was developed and validated by the researcher and aimed to measure the participants’ viewpoints toward the ChatGPT application. It consisted of 14 items assessing participants' perceptions of ChatGPT on a 5-point Likert scale, examining the respondents’ level of disagreement or agreement from 1 (strongly disagree) to 5 (strongly agree). The reliability of the scale was calculated to be 0.86 based on Cronbach's alpha with 125 EFL teachers. This number was greater than the total aggregate of the research participating teachers because during the selection process of the main participants, only 93 Iranian and non-Iranian English teachers expressed their willingness to participate in the research and were made available to the researchers. The content validity of the scale was also confirmed by five experts in TEFL and AI tools, relying on expert judgment validity. Construct validity of the scale was also confirmed through factor analysis, confirming the loadings of the variables on all the factors of the scale.

4. Findings

The analysis of the data collected from online questionnaires was conducted in two distinct sections, scrutinizing demographic information provided by participants and then examining the responses to the questionnaire items. The participants’ demographic information is presented in Table 1.

Table 1

Participants’ Demographic Characteristics

Comparing Iranian and non-Iranian English teachers, it is evident that there are some notable differences in gender distribution and academic qualifications. Iranian teachers had a significantly higher percentage of male teachers in comparison to non-Iranian teachers (91.1% vs. 72.9%). Additionally, while the majority of both Iranian and non-Iranian English teachers held a Master's degree, a slightly higher percentage of non-Iranian teachers held a PhD (10.4% vs. 4.4%). However, both groups reported similar years of teaching experience, with the largest proportion falling within the 11-20 years of experience range.

Moreover, descriptive statistics and the one-way ANOVA test were run to address the three questions of this study. The frequencies (F), relative frequencies (%), and means (M) were used to show the degree of both groups of teachers’ perspectives toward employing ChatGPT in English language instruction and assessment. Additionally, the one-way ANOVA test was conducted to examine whether there were significant differences between Iranian and non-Iranian English teachers’ perceptions regarding various aspects of ChatGPT application in English language assessment and teaching. As it was mentioned, the questionnaire consisted of 14 items examining different ChatGPT aspects such as knowledge of ChatGPT, ChatGPT effectiveness, and ChatGPT-related challenges. These items were analyzed in the following tables.  To answer the first question of understanding how both Iranian and non-Iranian English teachers perceive ChatGPT, an analysis of six key items was conducted. In this regard, descriptive statistics and the one-way ANOVA test were run. The result is illustrated in Tables 2 and 3.

Table 2

Descriptive Statistics of English Teachers’ Knowledge of ChatGPT

It is worth mentioning that for the cumulative estimation of English teachers' knowledge about ChatGPT, responses from the Likert scales 'Strongly Disagree' and 'Disagree' were classified together, as were responses from 'Strongly Agree' and 'Agree.' This consolidation allows for a comprehensive assessment of teachers' knowledge, providing insights into the overall agreement or disagreement trends across various aspects of ChatGPT.

As Table 2 illustrates, for the statement that "ChatGPT is a search engine like Google," the highest level of disagreement comes from both Iranian and non-Iranian English teachers, with 24 Iranian teachers (53.3%) and 26 non-Iranian English teachers (54.2%) disagreeing. This suggests a predominant disagreement among both sets of teachers regarding ChatGPT's resemblance to Google's search engine. It means that both groups of teachers have a good understanding that ChatGPT is not a search engine similar to Google. When evaluating "ChatGPT as a language model being able to generate content based on basic information," a notable level of agreement is evident, particularly among non-Iranian teachers. Specifically, 62.2% of Iranian teachers and 89.6% of non-Iranian teachers agree with this statement. It is necessary to mention that while ChatGPT can generate texts based on input, including keywords, its capabilities extend beyond basic keywords. ChatGPT can generate text that goes beyond simple keyword prompts and can understand the context to some extent. Therefore, the high agreement among both Iranian and non-Iranian teachers suggests a recognition of ChatGPT's broader language generation capabilities. Additionally, in terms of the perception that "ChatGPT is generated from a large and regularly updated database," both Iranian and non-Iranian teachers show substantial agreement, with 88.9% of Iranian teachers (n=40) and 85.4% of non-Iranian teachers (n=41) in agreement. It is worth noting that while ChatGPT was indeed trained on a large dataset, it is essential to clarify that its knowledge has not been constantly updated since its last training data in September 2021. Consequently, there might be limitations in terms of having the most up-to-date information.

Despite this, the high agreement among both groups suggests a widespread belief in the comprehensiveness and continuous updating of ChatGPT's database, albeit with some room for clarification regarding the actual updating frequency. Furthermore, in the assessment of "ChatGPT's ability to understand human requests and execute them, both Iranian and non-Iranian English teachers demonstrate similar levels of agreement. Specifically, 26 Iranian teachers (57.8%) and 26 non-Iranian teachers (54.2%) agreed. It is important to recognize that while ChatGPT can indeed understand and respond to human demands based on its training data and capacities, it may sometimes respond imperfectly. Moreover, complex requests might require additional context clarification to generate relevant and coherent responses. Therefore, the observed agreement among both groups suggests a shared understanding of ChatGPT's capabilities and limitations in processing human queries and providing responses.

Moreover, in evaluating the role of "ChatGPT as a translation support tool," it is evident that both Iranian and non-Iranian teachers exhibit significant levels of agreement. Among non-Iranian teachers, 24 (50%) express agreement with this role, while a larger proportion of Iranian teachers, specifically 35 (77.8%), also agree. This alignment in perception suggests that both groups recognize ChatGPT's capacity to aid in translation tasks by generating text in multiple languages. However, it is important to mention that ChatGPT is not explicitly designed as a dedicated translation tool, unlike some other specialized software options available. Lastly, in examining the perception of "ChatGPT as an online platform for learning foreign languages," Iranian teachers exhibit the highest level of agreement, with 27 (60%) agreeing. However, non-Iranian teachers show considerably lower agreement, with only 9 (18.8%) expressing support for this statement. Additionally, there is a notable disparity in disagreement, with 23 (47.9%) non-Iranian teachers indicating disagreement with the notion that ChatGPT serves as an online language learning platform. It suggests that non-Iranian teachers generally made accurate assessments, while it appears that Iranian teachers have a lesser understanding of this issue. This discrepancy in perceptions between Iranian and non-Iranian teachers may stem from differences in their understanding of ChatGPT's capabilities and its suitability for language learning objectives.

According to Table 3, the one-way ANOVA tests revealed that there were no significant differences in the knowledge of English teachers from Iran and other countries regarding certain aspects of ChatGPT. Specifically, for the aspects related to ChatGPT being a search engine similar to Google (F = 0.137, Sig. = 0.713), created from a large and regularly updated database (F= 0.084, Sig. = 0.773), and capable of understanding human requests and executing them accurately (F = 0.002, Sig. = 0.962), the analyses indicated no statistically significant discrepancies in perception. These findings suggest a consensus among respondents from various countries regarding these particular attributes of ChatGPT.

Table 3

Results of One-way ANOVA Test Regarding English Teachers’ Knowledge of ChatGPT

However, the one-way ANOVA analyses unveiled noteworthy distinctions in how Iranian and non-Iranian teachers are familiar with ChatGPT in specific aspects. Specifically, for the aspects related to ChatGPT being a language model being able to generate content based on basic information (F = 7.019, Sig. = 0.010), a translation support tool (F = 6.283, Sig. = 0.014), and an online platform for learning foreign languages (F = 20.655, Sig. = 0.000), the analyses indicated statistically significant discrepancies in perception. These findings suggest varying perspectives among respondents from different countries regarding these particular attributes of ChatGPT, indicating a need for further investigation into the underlying reasons for such differences. Moreover, question two addressed assessing the effectiveness of ChatGPT in language instruction and assessment, as perceived by Iranian and non-Iranian English teachers, yielded notable findings as shown in Tables 4 and 5.

Table 4

Descriptive Statistics of English Teachers’ Perceptions of ChatGPT Usefulness

According to Table 4, for the statement "I find ChatGPT a helpful tool in language classes, encouraging learners to learn," the highest frequency of positive responses comes from Iranian teachers, with 17 (37.8%) strongly agreeing. Non-Iranian teachers also view it favorably, with 19 (39.6%) agreeing that ChatGPT is motivational. Moreover, regarding the belief that "ChatGPT can save time in supplying feedback and grading for learners," both Iranian and non-Iranian teachers strongly agree and agree, with 17 (37.8%) and 17 (35.4%) respectively, suggesting that ChatGPT is seen as a time-efficient tool. However, the non-Iranian teachers show a more varied response, with a significant 14 (29.2%) remaining neutral. When considering the statement "I am more confident in my activity when I have support from ChatGPT," a notable 20 Iranian teachers (44.4%) agree, reflecting a reliance on ChatGPT for professional confidence. Non-Iranian teachers also express confidence, with 18 (37.5%) agreeing, though the strong agreement is less pronounced at 5 (10.4%). Finally, for "I find ChatGPT to be a valuable self-learning assistant for learners," the highest frequency of agreement is among non-Iranian teachers, with 20 (41.7%) agreeing. Iranian teachers also recognize its value, with 15 (33.3%) strongly agreeing. This suggests that ChatGPT is considered a significant asset for student-led learning.

Table 5

Results of One-way ANOVA Test About English Teachers’ Perceptions of ChatGPT Usefulness

As illustrated in Table 5, significant differences were observed between Iranian and non-Iranian teachers in their perceptions of ChatGPT's utility in the classroom for encouraging students (F = 9.396, Sig. = 0.003), saving time in providing feedback and grading (F = 7.807, Sig. = 0.006), and instilling confidence in their work with ChatGPT's support (F = 10.716, Sig. = 0.002). However, no statistically significant difference was found in the perception of ChatGPT as a valuable self-learning assistant for language learners between Iranian and non-Iranian teachers (F = 2.059, Sig. = 0.155). To address the third question, descriptive statistics and the one-way ANOVA tests were conducted to examine the perceptions of Iranian and non-Iranian English teachers regarding the challenges of employing ChatGPT in language instruction and assessment (See Tables 6 and 7). The data from Table 6 sheds light on English teachers' perceptions of the challenges of ChatGPT application in educational settings. A significant number of Iranian teachers view ChatGPT positively, with 29 (64.4%) either disagreeing or strongly disagreeing that it is unhelpful and a time-waster. Non-Iranian teachers share this sentiment, with 27 (56.3%) also disagreeing or strongly disagreeing with the statement. This indicates a general belief that ChatGPT is not seen as detrimental to classroom efficiency. Moreover, concerns about academic integrity are evident, as a majority of both Iranian and non-Iranian teachers agree or strongly agree that ChatGPT could facilitate cheating on assignments and exams, with 39 (86.7%) and 36 (75%) respectively. This reflects a significant apprehension regarding the potential misuse of the tool by students.

Table 6

Descriptive Statistics of English Teachers’ Perceptions of ChatGPT Challenges

Additionally, skepticism about the accuracy of information supplied by ChatGPT is also present, but less pronounced. Among Iranian teachers, 18 (40%) agree that they are skeptical, while 17 (35.4%) non-Iranian teachers share this skepticism. However, a considerable proportion remains neutral, suggesting uncertainty or a lack of firm opinions on the matter. Lastly, the concern that students may become less inclined to think critically if they rely too much on ChatGPT is shared by both groups, with 30 (66.7%) Iranian teachers and 33 (68.7%) non-Iranian teachers agreeing or strongly agreeing. This suggests a common belief that excessive use of ChatGPT could potentially hinder students' cognitive engagement.

Table 7

Results of One-way ANOVA Test About English Teachers’ Perceptions of ChatGPT Challenges

In terms of finding ChatGPT to be unsupportive in language classes and spins learners’ wheels, no significant difference was observed between the perceptions of Iranian and non-Iranian teachers (F = 0.625, Sig. = 0.431). Similarly, regarding the belief that ChatGPT can provide the opportunity for students to cheat on assignments and/or exams, there was no significant disparity between Iranian and non-Iranian teachers' viewpoints (F = 1.815, Sig. = 0.181). Additionally, concerning skepticism about the accuracy of the information supplied by ChatGPT, no significant difference was found between Iranian and non-Iranian English teachers (F = 0.416, Sig. = 0.521). Furthermore, regarding concerns about students becoming lazier because of using ChatGPT continuously, the analysis revealed no significant difference between Iranian and non-Iranian teachers' perceptions (F = 0.400, Sig. = 0.529).

5. Discussion

This research paper compared the perceptions of both Iranian and non-Iranian teachers on the applicability and limitations of using ChatGPT in language teaching and assessment. Hence, the degree of the perception of each group toward the ChatGPT application in English language teaching and assessment was primarily investigated, followed by providing insights into the Iranian and non-Iranian English teachers’ perceptions concerning different dimensions of ChatGPT using the one-way ANOVA test results. The findings revealed that both Iranian and non-Iranian teachers recognize that ChatGPT differs from traditional search engines like Google, acknowledging its unique text-generation abilities. However, there may be some overestimation of its updating capabilities, highlighting the need for better awareness of its limitations. While there is agreement on ChatGPT's ability to understand and execute human requests, there is also an acknowledgment that its responses may not always be perfect. Similarly, there is recognition of ChatGPT's role as a translation support tool, although it is not perceived as a specialized translation tool. One notable difference emerges regarding ChatGPT's suitability as an online platform for learning foreign languages. Iranian teachers show higher agreement, while non-Iranian teachers exhibit more skepticism, indicating a more critical assessment of its language learning capabilities. These findings are consistent with Kiryakova & Angelova (2023). However, the results are different from those of Polak et al. (2022), indicating that although teachers have basic digital skills, they generally hold low AI-related skills. In sum, while English language teachers generally grasp ChatGPT's functionalities, there are areas where more clarity is needed, particularly regarding its updating mechanism and suitability for language learning. This highlights the importance of providing educators with comprehensive information about AI tools like ChatGPT to ensure their effective integration into educational settings.

Furthermore, Iranian teachers, in particular, see it as a motivational tool, with a significant number strongly agreeing. Both Iranian and non-Iranian teachers acknowledge its time-saving benefits for grading and feedback, though non-Iranian teachers show more neutrality. Confidence in work is notably higher among Iranian teachers when supported by ChatGPT. Additionally, non-Iranian teachers see ChatGPT as a valuable self-learning assistant for learners. These findings highlight ChatGPT's potential to motivate and encourage students, save time for teachers, boost teacher confidence, and facilitate student-led learning. However, careful integration and ongoing evaluation are crucial to maximize its benefits while addressing any challenges. Similarly, Ali et al. (2023) revealed that ChatGPT is an encouraging tool for instruction.

While most teachers, regardless of nationality, view ChatGPT positively, concerns over academic integrity and the reliability of information are prevalent. This finding is in line with previous studies (Kiryakova & Angelova,2023; Zargoun, 2024), which concluded that teachers express caution about potential effects on students' critical thinking abilities, highlighting the significance of a balanced way to AI integration. Overall, the data emphasized the need for careful consideration of AI's role in education, ensuring it enhances rather than undermines traditional teaching methods and critical thinking development. Moreover, it is noteworthy that no significant differences were found in their knowledge of certain fundamental features of ChatGPT, such as its resemblance to a search engine, its data source, and its ability to understand and execute human requests accurately. This suggests a uniform understanding and perception of these basic functionalities across different cultural backgrounds. However, the analyses also revealed significant distinctions in the perception of ChatGPT's capabilities as a language model, translation support tool, and online learning platform between Iranian and non-Iranian teachers. These differences imply that while there may be a consensus on basic features, there are divergent views on the more nuanced aspects of ChatGPT's utility in language learning and teaching. For instance, the significant difference in perception regarding ChatGPT as a language model that can generate content based on basic information suggests that Iranian teachers may have a different understanding or experience with this aspect compared to their non-Iranian counterparts. Similarly, the disparities in perception regarding ChatGPT as a translation support tool and an online learning platform for foreign languages indicate varying levels of acceptance or skepticism toward these functionalities among teachers from different cultural backgrounds.

Secondly, significant differences were observed in how these two groups of teachers perceive ChatGPT's effectiveness in motivating students, saving time in providing feedback and grading, and instilling confidence in their work with ChatGPT's support. The significant disparities in these areas suggest that Iranian and non-Iranian teachers may have distinct views on the practical benefits and challenges associated with integrating ChatGPT into their teaching practices. For instance, the finding that non-Iranian teachers perceive ChatGPT as more effective in motivating and encouraging students and saving time could indicate a greater familiarity or comfort with technology-enhanced language teaching methods compared to their Iranian counterparts. On the other hand, Iranian teachers may have concerns about the effect of ChatGPT on student motivation and academic integrity, leading to a more cautious approach toward its adoption in the classroom. Furthermore, the significant difference in perception regarding confidence in work with ChatGPT's support suggests that Iranian and non-Iranian teachers may have differing levels of trust in the accuracy and reliability of ChatGPT-generated content. This could stem from differences in educational backgrounds, cultural attitudes toward technology, or previous experiences with AI-powered tools in education. However, it is worth noting that no statistically significant difference was found in the perception of ChatGPT as a powerful self-learning assistant for learners between Iranian and non-Iranian teachers. This indicates a consensus among both groups regarding ChatGPT's potential to support student learning independently.

Thirdly, the results showed that Iranian and non-Iranian teachers did not have significant differences in their perceptions of the possible disadvantages and problems associated with the ChatGPT application in the classroom setting. Importantly, the two groups were not statistically significantly different in their perceptions of the unsupportive and potentially time-wasting nature of ChatGPT, on the one hand, and the facilitating role it had in student cheating on assignments and exams, on the other hand. It was also found that the two groups were not significantly different in terms of skepticism about the reliability of ChatGPT-created information and concerns that excessively using ChatGPT would adversely affect the students’ thinking, making them lazier. The findings of the present study agree with the results reported by Baidoo-Anu and Owusu Ansah (2023), highlighting some teachers’ skepticism toward AI and regarding it as a hazard that would potentially end most educational activities, making instructors and learners lazy and giving them no or little analytical capabilities.

Moreover, these findings suggest a consensus among both Iranian and non-Iranian teachers regarding the perceived drawbacks and challenges associated with ChatGPT. Both groups may share similar concerns and reservations about the utilization of ChatGPT in educational settings, regardless of their cultural or geographical backgrounds. One possible explanation for the lack of significant differences in perception could be the universality of concerns related to the use of AI technologies in education. Issues such as reliability, academic integrity, and the potential effect on student learning and engagement are likely to be recognized and discussed across different cultural contexts and educational systems. Therefore, Iranian and non-Iranian teachers may approach these concerns with similar levels of caution and skepticism, leading to consistent perceptions across both Iranian and non-Iranian English teachers. Furthermore, it is essential to consider the overall familiarity and experience of both Iranian and non-Iranian teachers with AI technologies like ChatGPT. Regardless of cultural differences, teachers may share common apprehensions about the potential challenges and limitations associated with integrating AI tools into their teaching practices, which could contribute to the alignment in their perceptions observed in this study.

6. Conclusion

           Generative AI (GenAI) refers to AI systems that can generate high-quality content in the form of text, images, and others, based on what they have been trained on (Martineau, 2023). GenAI is differentiated from traditional AI, which employs prediction algorithms and historical data, in that it generates varied outputs in the forms of text, code, images, videos, and biological entities such as molecules (Locky Law, 2024). Of these, chatbots are especially excellent examples of GenAI. Sometimes referred to as conversational agents, chatbots are advanced computer systems that interact with humans through interfaces such as messaging applications (Ciechanowski, 2019). Advancements in AI and Natural Language Processing (NLP) have enabled professional and context-aware conversations with chatbots. The systems are now able to process vast amounts of data to determine conversational context and user intent, and perform tasks such as personalized advice, deep question answering, and user-specific tuning (Liu et al., 2022). ChatGPT has gained significant attention as one of the highest-ranked generative AI chatbots (Chen & Lin, 2020; Chiu, 2023; Hsu et al., 2023). The availability and simplicity of AI-powered tools have resulted in their extensive application by teachers and learners (Alharbi, 2023). In English language learning, conversational AI tools are employed to provide interactive and immersive learning environments. The tools enhance the students’ level of confidence to speak (Ruan et al., 2021), encourage them to learn (Liu et al., 2022), and provide them with personalized feedback to improve their writing (Hwang et al., 2023).

The current paper focuses on providing a comparative analysis of Iranian and non-Iranian EFL teachers' perceptions of the applications of ChatGPT and its related challenges in language instruction and assessment. Drawing on the results, there is a consensus on the utility of ChatGPT, but there are divergent standpoints about more advanced applications of this technology. Iranian teachers are more favorable in their evaluation of the utility of ChatGPT in language teaching compared to their non-Iranian counterparts, while non-Iranian teachers are more cautious. All the participants also highlight the utility of ChatGPT in grading and giving feedback. Concerns about academic integrity and the authenticity of AI-generated information, however, point to the necessity to have a balanced approach to incorporating such tools in academic settings. As shown by the results, the participants appreciate ChatGPT's text generation capabilities, recognizing its unique capabilities beyond conventionally used search engines. Nevertheless, the significant limitations of ChatGPT should not be overlooked, especially regarding its updating abilities and contribution as a specialized translation assistant. It was also highlighted through this study that, despite the consensus on ChatGPT’s basic practical applications, its advanced characteristics and feasibility in educational contexts are mainly questioned. Unlike non-Iranian teachers who look at ChatGPT's contribution to language learning from a more critical standpoint, Iranian teachers approach it more optimistically.

Significant differences were also observed in the perceived effectiveness of ChatGPT in motivating students, saving time for grading and feedback, and instilling confidence in teachers' work. These disparities suggest varying degrees of acceptance and skepticism toward ChatGPT's integration into teaching practices across different cultural backgrounds. Additionally, despite these differences, there is a shared concern among all teachers about the potential effect of ChatGPT on academic integrity and the reliability of information. This reveals the significance of a balanced strategy for AI integration that supports traditional teaching methods and fosters the development of critical thinking skills.

Similar to other studies, the current research suffered from some limitations, although it provided insights into the participants’ perceptions of the contributions of ChatGPT to language teaching and assessment. First, the accuracy of the teachers' perceptions might be potentially affected by subjective biases due to reliance on self-reported data. Another limitation is attributed to the small sample size, potentially affecting the generalizability of the results, as the sample may fail to fully represent the broader population. In addition, due to the increasingly fast growth of AI technologies, more comprehensive studies are needed to emphasize the necessity of ongoing research to keep up with technological developments. It should also be noted that researchers need more caution when interpreting the results and seeking to strengthen research foundations for broader contexts with higher diversity. Hence, the long-term effects of ChatGPT on language learning and teacher pedagogy should be investigated through further research while also examining the particular training requirements of teachers for effective ChatGPT integration into their teaching strategies. Researchers are also recommended to conduct comparative research across various educational and cultural settings to provide a deeper understanding of the applicability of ChatGPT to language instruction. Overall, the current study sought to play an effective role in the rapidly evolving knowledge of AI in education while highlighting that tools like ChatGPT should be thoughtfully integrated into education to ensure enhanced language teaching and learning experiences. In line with these goals, educators require comprehensive knowledge and awareness of AI capabilities and restrictions to promote its full potential while dealing with the existing challenges.