Computers and Education Open 6 (2024) 100174

Contents lists available at ScienceDirect

Computers and Education Open

journal homepage: www.sciencedirect.com/journal/computers-and-education-open

Application of generative artificial intelligence (GenAI) in language

teaching and learning: A scoping literature review
Locky Law (Locky lok-hei Law)*
Centre for Applied English Studies, The University of Hong Kong

A R T I C L E I N F O A B S T R A C T

Keywords: This scoping literature review examines the application of Generative Artificial Intelligence (GenAI), a disruptive
Generative AI technology, in language teaching and learning. Since its launch in November 2022, GenAI has captured global
AI attention with OpenAI’s ChatGPT, powered by the generative pre-trained transformer-3 (GPT-3) large-language
Language education
model. The emergence of GenAI holds immense implications across various domains, including language edu­
Scoping review
Content generation
cation. This review aims to provide an overview of the current state of research and identify research gaps and
ChatGPT future directions in this emerging field. The review follows the PRISMA-ScR guidelines and includes eligible
publications published between 2017 and July 2023. Four electronic databases were searched and 41 of the 224
initial papers were eventually selected for review. The findings reveal key terms related to GenAI in language
education, the most researched language study and education levels, areas of research, attitudes towards GenAI,
and the potential benefits and challenges of GenAI application. The review highlights several research gaps,
including the need for more empirical studies to assess the effectiveness and impact of GenAI tools, discussion of
ethical considerations, targeted interventions for specific language skills, and stakeholder engagement in
responsible integration. Educators are encouraged to incorporate GenAI tools into their teaching practices while
remaining vigilant about potential risks. Continuous professional development for educators is crucial to ensure
informed decision-making and effective integration of GenAI tools. This scoping review contributes to the
existing knowledge on the use of GenAI in language education and informs future research and practice in this
disruptive and rapidly evolving field.

1. Introductions pre-trained transformer (GPT), ChatGPT has demonstrated the ability to

generate coherent and grammatically correct text while continuously
Generative artificial intelligence (GenAI), according to [9] defini­ improving through machine learning from user inputs. The emergence
tion, is “the use of AI to create new content, like text, images, music, of GenAI, particularly exemplified by ChatGPT, holds immense impli­
audio, and videos”, utilizing a machine learning (ML) model “to learn cations across various domains, including language teaching and
the patterns and relationships in a dataset of human-created content” learning.
and then “uses the learned patterns to generate new content.” As a Language teaching and learning play a vital role in today’s global­
subset of artificial intelligence (AI), GenAI differs from past forms of AI ized world, where effective communication and intercultural under­
technology that employ ML algorithms and predication from data based standing are essential for personal, academic, and professional success.
on past behavior. Instead, according to [49], GenAI focuses on creating Proficiency in a language enables individuals to engage in meaningful
new textual and multimodal content using large language models interactions, express clear ideas, and navigate diverse cultural contexts.
(LLMs), art-based models and video-based models. Examples of Traditionally in the academic setting, language educators have been the
well-known GenAI tools include OpenAI’s ChatGPT, GPT-4, Playground, primary facilitators or catalysts for language acquisition and develop­
DALL ⋅ E 3, and Sora, Anthropic’s Claude, Google’s Gemini (previously ment, where students learn essential language skills from teachers who
Bard), Stability AI’s Stable Diffusion 3, and Runaway’s Gen-2. assess their progress through written and spoken components. However,
Since the launch of [32] ChatGPT in November 2022, GenAI has the advent of the internet and search engines has transformed the lan­
attracted global attention. Powered by a LLM known as generative guage learning landscape, as it drastically reduces the reliance of

* Corresponding author at: Room 1307, Block P, Kornhill, Quarry Bay, Hong Kong.
E-mail address: lockylaw@hku.hk.

https://doi.org/10.1016/j.caeo.2024.100174
Received 16 December 2023; Received in revised form 13 March 2024; Accepted 27 March 2024
Available online 28 March 2024
2666-5573/© 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
L. Law Computers and Education Open 6 (2024) 100174

students on teachers, allowing students to access vast amounts of in­ 2. Methods

formation, language resources, and language learning platforms that
cater to their individual learning needs [43]. The introduction of GenAI 2.1. Exploratory literature review with PRISMA-ScR
programs represents the beginning of yet another paradigm shift in
language education and acquisition. Given GenAI has only started gaining traction in the research field
At this early stage of GenAI development, conducting a scoping re­ near the end of November 2022 despite years of development, there is
view becomes crucial and highly valuable, as it provides an overview of inadequate literature to support a systematic approach to literature re­
the current research landscape and holds implications for major stake­ view. This makes a scoping literature review a more suitable approach
holders involved in language education. By examining the existing [45]. Also, compared to systematic reviews, scoping reviews are “useful
literature, the review can shed light on the potential impact of GenAI on for answering much broader questions (such as “What is the nature of
language teaching and learning, inform future research directions, and the evidence for this intervention?” or “What is known about this
guide the actions of language educators, policymakers, developers, and concept?”)” ( [45], p. 467). Because of the difference in the objectives of
researchers in this evolving field. this study, I adopted the PRISMA statement on scoping review (PRIS­
MA-ScR) by [45].

1.1. Rationale 2.2. Eligibility criteria and information sources

While there is a substantial body of published studies and literature The sources of evidence are considered eligible based on years of
reviews on artificial intelligence (AI) in recent years (e.g., [6,21,53]), publication, language and publication type. Publications must be pub­
the existing literature specifically focusing on the application of GenAI lished between 2017 and 2023, from the year the Transformer (the ‘T’ in
in language education, as well as GenAI in general, is still very limited. GPT, generative pre-trained transformer) was first announced [47] up
Given that GenAI is a relatively new and rapidly evolving technology, a till the time of this writing on July 25, 2023. The idea is to capture the
scoping review presents a suitable methodology for providing a recent development in GenAI in recent years as well as any information
comprehensive overview of the current state of research in this area. The related to its build-up a decade prior to this. Also, this study only
advantage of a scoping review lies in its ability to encompass diverse considered only English papers as it is the only language the author is
study designs, methodologies, and types of literature, including empir­ proficient in. Eligible publication types include published theoretical
ical studies, technology reviews, and review articles. By undertaking a paper, empirical study, review article, technology review, editorial
scoping review that examines the application of GenAI in language opinion, and discussion papers. Gray literature is also eligible as it
teaching and learning, researchers can gain valuable insights into the valuable in shaping research direction at the early stage of GenAI
current research landscape and identify areas that warrant further technological development. However, book reviews were not included
investigation. This process facilitates the identification of research gaps as they are not complete studies on a topic.
and emerging trends, thereby guiding future research endeavours in this Several exclusion criteria were considered in the screening phase. I
domain. Furthermore, the findings derived from the scoping review also exclude publications that do not have a language education focus or
have practical implications as they inform the development of research GenAI focus because these are not within the scope of this study. Pre­
questions and support evidence-based decision-making in the field of prints of published work which are already in the record were also
language education with GenAI. Notably, to the best of my knowledge, removed to avoid duplication of content. If for any reasons that the
no scoping literature review has been conducted thus far, specifically abstract of a publication was not retrievable, the record was also
focusing on the application of GenAI in language education. Conse­ excluded. Finally, this study does not include students’ assignments,
quently, this research represents a pioneering effort to address this commercial websites, blogs, magazine articles, conference abstract as
critical gap in the literature. these text types are not intended to be or published by academic
journals.
1.2. Objectives
2.3. Search and selection of sources of evidence
To conduct a scoping review of the literature to determine the cur­
rent state of research on the use of GenAI (both textual and multimodal) The search was conducted in four electronic databases: SCOPUS
in language teaching and learning, and to identify research gaps and (Elsevier), Science Direct (Elsevier), JSTOR (ITHAKA), ERIC (Institute of
areas for future investigation. Education Sciences), and Google Scholar (Alphabet). Given that GenAI is
a relatively new topic in language education, gray literature was
considered for screening.
1.3. Review questions Potential literature was identified using the Boolean conjunction:
("generative artificial intelligence" OR "generative AI" OR "GenAI" OR
This scoping literature review addresses the following review "ChatGPT" OR “Chat GPT”) AND ("language education" Or "language
questions: teaching" OR "language learning"). I included studies that focused on the
use of GenAI in language teaching and learning, irrespective of the
1. What are the key terms surrounding the use of GenAI in language language or level of proficiency.
education? The process for selecting sources of evidence included in this scoping
2. What language studies and education levels are most researched? review includes identification, and screening and eligibility.
3. a) What are the specific areas of research (i.e., language learning/ Adapting the design of tables from [21] scoping review paper, new
language skills that have been investigated) in relation to the use of tables were constructed to meet the needs of this study and were filled
GenAI? out carefully after close reading the selected literature several times. I
also used AntConc (v 4.2.2) [4] to build a corpus using the selected
b) What are the attitudes towards the use of GenAI in language literature which allowed for the queries of keywords, concordances, and
teaching and learning? frequency of occurrences, which in turn facilitated the search for
important information and confirmation of the accuracy of my data
1. What are the potential benefits and challenges of using GenAI in extraction. Re-reading of the papers were performed, and adjustments
language teaching and learning? were made to the cell content when necessary.

2
L. Law Computers and Education Open 6 (2024) 100174

3. Results entry was published in 2022 [44], and all others in 2023. This is not
surprising because GenAI in language education mostly began to gain
3.1. Selection of sources of evidence attention since the launch of ChatGPT in Nov 2022.
Among the 41 selected papers (100 %), 33 are journal articles (80
The flow diagram showing the selection of sources of evidence is %), 7 are gray literature (17 %), and 1 editorial position paper (2 %).
summarised in Fig. 1. In the identification phase, my search yielded an Study types consist of 24 empirical studies (59 %) (i.e., 9 qualitative, 8
initial 224 entries (100 %) from selected databases and no additional mixed methods, and 7 quantitative studies), 6 review articles (15 %), 10
records were identified through other sources. The citation information technology reviews (24 %), and 1 commentary (2 %) (see Table 1).
of these entries was downloaded from the databases and stored in a Regarding the origins of the papers, I only considered empirical
Microsoft Excel spreadsheet which was used to remove duplicates. 195 studies because origins of other study types are not as relevant, such as
records (87 %) remained after duplicates were removed. These records review articles and technology reviews. It is found that East Asia regions
were screened, and 117 records were excluded, leaving 78 records (35 (Hong Kong, China, South Korea, Japan) are most productive with ten
%) for full-text article retrieval and assessment for eligibility. A total of studies (42 %) [10,14,18,23,40,50–52,55,56], followed by Middle East
37 papers were excluded due to reasons such as failure to retrieve full regions (Iraqi Kurdistan, Saudi Arabia, Turkey) with three (13 %) [2,29,
text, absence of a GenAI focus, and poor quality (e.g., missing major 30], two (8 %) from Southeast Asia (Indonesia) ([1]; [46]) and one each
sections, such as methodology). In the end, a total of 41 studies (18 %) from Benin [58], Chile [48], and Czech Republic [17]. However, three
that met the inclusion criteria. empirical studies have not specified the data origins (17 %) [3,27,31],
while three others used internet data of which the origins cannot be
identified (13 %) [24,25,28].
3.2. Characteristics of sources of evidence

While the search focused on identifying records published between

2017 and 2023, up till the time of this writing on July 25, 2023, only one

Fig. 1. A flow diagram summary of the selection of sources of evidence.

3
L. Law Computers and Education Open 6 (2024) 100174

Table 1 publications either have not specified any education levels or that the
Characteristics of sources of evidence in frequency and percentages. education level is not applicable in the discussion.
Characteristics Frequency %
3.3.3. Review question 3: a) areas of research and b) attitudes towards
Publication year 41 100
2022 1 2 GenAI
2023 40 98 Within the scope of language teaching and learning (T&L), there are
Publication type 41 100 several major areas of research documented in the forty-one selected
Journal article 33 80 literature. The predominant area of investigation pertains to general
Gray literature 7 17
Editorial position paper 1 2
T&L, which encompasses a significant portion of the literature (22
Study type 41 100 studies). Following this, T&L policy has garnered substantial attention,
Empirical study 24 59 with nine studies dedicated to this topic ([1,3,10,18,19,23,27,38,39]).
Qualitative 9 22 The domains of writing and assessment have also received notable
Mixed methods 8 20
scholarly interest, with five studies focusing on the former [16,25,40,46,
Quantitative 7 17
Review article 6 15 56] and two studies investigating the latter [20,28]. The discussion on
Technology review 10 24 ethics, which is closely related to T&L policy, also yield two studies [35,
Commentary 1 2 36]. However, only a single study has examined a specific topic and it is
Origin of empirical study (by region) 24 100 on fallacy learning [55].
East Asia 10 42
Hong Kong 4 17
Among the selected papers, the most widely studied application of
China 3 13 GenAI in language teaching and learning has been its use for writing
South Korea 2 8 instruction. Studies have demonstrated that GenAI systems can, to
Japan 1 4 various extents, assist learners improve their writing skills by providing
Middle East 3 13
real-time feedback on grammar, vocabulary, and sentence structure ([1,
Iraqi Kurdistan 1 4
Saudi Arabia 1 4 40,52]). These systems can also help learners expand their vocabulary
Turkey 1 4 and improve their sentence structures by providing suggestions for
Southeast Asia 2 8 alternative word choices and sentence rephrasings [38]. However, some
Indonesia 2 8 papers have also pointed out the limitations of GenAI in writing in­
Other (Benin, Chile, Czech Republic) Each 1 13
struction, such as the potential for overreliance on the system’s sug­
Unspecified 3 13
Internet data 3 13 gestions, may hinder learners’ critical thinking skills [27,29,30].
Regarding the utilization of GenAI in T&L, thirty-six papers express a
clear attitude towards the matter. Among these, twenty-eight papers
3.3. Results of individual sources of evidence exhibit a positive attitude towards GenAI implementation, four papers
express a mixed attitude [29,30,50,56], three papers strive to present a
Table 2 and Table 3 provide a summary of information on various balanced perspective [12,14,57], while only one paper adopts a nega­
aspects related to the use of GenAI programs in language teaching and tive attitude [52]. In summary, the primary area of application for
learning. The concepts and themes that emerged from the analysis GenAI tools is language T&L. The studies emphasize the value of
include research types and methods, subject locations/origins, educa­ incorporating GenAI in language classrooms to support language
tion levels, areas of research, attitudes, language skills, testing, assess­ acquisition, improve language skills, and offer personalized learning
ments, materials development, and policies. These themes encompass a experiences. Researchers highlight the potential of GenAI tools, such as
broad range of studies that trial GenAI in language education, high­ ChatGPT, in enhancing various aspects of education, including language
lighting the breadth of research conducted in this area. The results skills, content generation, personalized learning, and assessment.
presented in the tables directly address the objective and review ques­
tions established for this scoping review. 3.3.4. Review question 4: Potential impacts of GenAI (opportunities and
challenges)
3.3.1. Review question 1: key terms in genai in education/language Researchers have suggested the potential positive impacts of GenAI
learning writing tools such as ChatGPT on L2 writing and on certain psycholog­
In the literature, GenAI is commonly described as a state-of-the-art ical aspects such as learning motivations [10,29,31], interest and
[11,16,19,25,28,52] AI-powered large language model (LLM) technol­ engagement [31] and students’ writing creativity [27]. In the context of
ogy [14,35] existing in the form of Natural Language Processing (NLP) language teaching and learning, many studies have suggested that
software tools [e.g., chatbots [55]]. When discussed under the context of GenAI (including ChatGPT and other chatbots) and LLMs have the po­
language education, the technology falls under the umbrella terms of tential to offer advantages such as personalized learning, rapid re­
AI-based learning tools [46], chatbot-based learning tools [55], AI sponses, improved language learning outcomes, and enhanced learning
dialog systems for EFL [54], as well as AI-powered platforms and ap­ experiences and autonomy. For example, [54] highlights the potential
plications [48]. Intriguingly, within the context of GenAI, the term “AI benefits of the use of AI dialog systems for EFL education, which in­
literacy” does not seem to be vastly salient in the selected literature, cludes customisable input and complexity and instant feedback. [58]
appearing only five times in total in only two articles [36,46]. argues that AI-powered collaborative and interactive language learning
tools can enhance student engagement in EFL teaching, improve student
3.3.2. Review question 2: English as a foreign language (EFL) and diverse learning outcomes, and increase teacher satisfaction. In addition, [1]
education levels argues that ChatGPT has the potential to promote autonomy and
Most publications center around English as a Foreign Language (EFL) personalized learning in language education via its human-like con­
learners with the exceptions of [46] who looked into Indonesian texts, as versation and tailored language learning assistance. Learners can even
well as [2] who also discuss the possibilities of GenAI for teaching social regulate their learning processes, set objectives, and make decisions
studies, maths, and Turkish. about their language acquisition. [57] further argue that ChatGPT can
The application of GenAI is explored across various educational be deployed as a virtual tutor/expert/learning buddy or simply as a
levels and international tests, including preschool [44], primary [2,14, reflective learning tool or a stimulus for critical thinking. In terms of
23,44], secondary ([1,46,50,51]), higher education [8,10,18,19,27,29, materials development and lesson planning, [20] demonstrates how
30,35,40,48,52,55,56,58], CEFR [5,20] and TOEFL [28]. The remaining ChatGPT can assist EFL teachers in efficiently generating assignments,

4
L. Law Computers and Education Open 6 (2024) 100174

Table 2
Selected papers by author, research type, subject location/origin, education level, area of research, and attitude.
Author Research type Subject location/ Education level Area of research Attitude
origin

[1] Empirical (mixed Indonesia Secondary T&L policy Positive

methods)
[2] Empirical (qualitative) Turkey Primary T&L Positive
[3] Empirical (quantitative) Not specified Unknown T&L policy Positive
[5] Technology review N.A. International English tests: T&L Positive
CEFR
[8] Technology review Hong Kong University T&L Positive
[10] Empirical (quantitative) China University T&L policy Positive
[11] Review article N.A. Not specified T&L Positive
[12] Technology review N.A. N.A. T&L Balanced
[13] Technology review N.A. N.A. T&L Positive
[15] Empirical (qualitative) South Korea Primary T&L Balanced
[14] Review article N.A. Not specified T&L Positive
[16] Review article N.A. N.A. Writing N.A.
[17] Empirical (quantitative) Czech Republic Not specified T&L N.A.
[18] Empirical (qualitative) Hong Kong University T&L policy Positive
[19] Technology review N.A. University T&L policy Positive
[20] Technology review N.A. International English tests: Assessment & Material Positive
CEFR development
[23] Empirical (mixed South Korea Primary T&L policy Positive
methods)
[24] Empirical (qualitative) Internet data N.A. T&L Positive
[25] Empirical (quantitative) Internet data N.A. Writing Positive
[27] Empirical (qualitative) Not specified University T&L policy Positive
[28] Empirical (quantitative) Internet data International English tests: Assessment: Essay grading Positive
TOEFL
[29] Empirical (qualitative) Saudi Arabia University T&L Mixed
[30] Empirical (mixed Iraqi Kurdistan University T&L Mixed
methods)
[31] Empirical (quantitative) Not specified Unknown T&L Positive
[35] Review article N.A. University Ethics N.A.
[36] Commentary N.A. N.A. Ethics N.A.
[38] Review article N.A. N.A. T&L policy N.A.
[39] Technology review N.A. Not specified T&L policy Positive
[40] Empirical (mixed Japan University Writing Positive
methods)
[42] Technology review N.A. Not specified T&L Positive
[44] Technology review N.A. Preschool and primary T&L Positive
[46] Empirical (mixed Indonesia Secondary Writing Positive (but not so good in
methods) Indonesian)
[48] Empirical (mixed Chile University T&L Positive
methods)
Woo et al. Empirical (qualitative) Hong Kong Secondary T&L Mixed
(2023)
Woo et al. Empirical (qualitative) Hong Kong Secondary T&L Positive
(2023)
[52] Empirical (qualitative) China University T&L Negative
[54] Review article N.A. N.A. T&L Positive
[55] Empirical (mixed Hong Kong University Fallacy learning Positive
methods)
[56] Empirical (quantitative) China University Writing Mixed
[57] Technology review N.A. Not specified T&L Balanced
[58] Empirical (mixed Benin, Africa University T&L Positive
methods)

quizzes, learning activities and lesson plans, saving valuable time and level of accuracy and reliability of scoring L2 writing and can be
effort which could be used on other important aspects of language enhanced by integrating linguistic features such as lexis, syntax, and
pedagogy. cohesion.
A few papers have discussed the potential of GenAI language models However, concerns have been raised regarding the potential chal­
in evaluating writing tasks, providing real-time feedback, and revolu­ lenges, limitations, and pedagogical impact of GenAI in the literature.
tionizing writing evaluation methods. For instance, [29] discusses how More than half the selected papers have mentioned the ethical consid­
ChatGPT can enhance conventional pedagogies and improve the effi­ erations regarding the use of GenAI, with keywords involving plagia­
ciency of EFL teachers in grading while providing more accurate and rism, and biases. [52] is perhaps the most explicit about the
insightful feedback. [20] suggests that using ChatGPT for grading can disadvantages of GenAI, arguing that the threats of ChatGPT outweigh
streamline the assessment process, saving time and effort for teachers. its potential benefits. Citing related literature and results from the au­
By automating certain aspects of grading, teachers can focus on other thor’s study, [52] lists out several key threats of ChatGPT, including
critical aspects of language pedagogy. Finally, [28] have studied academic dishonesty (i.e., students take credit of AI-generated writing),
GenAI-powered TOEFL essay scoring in their research, specifically educational inequality (i.e., the GenAI program may create disparities
Automated Essay Scoring (AES) using GPT (Generative Pre-trained among students due to unequal accessibility to and proficiency in using
Transformer). Results show that AES by GPT can achieve a certain the tool), plagiarism detection avoidance (i.e., ChatGPT produces

5
L. Law Computers and Education Open 6 (2024) 100174

Table 3 Table 3 (continued )

Selected papers by author, data/participants, instrument, and summary. Author Data/Participants Instrument Summary
Author Data/Participants Instrument Summary
understanding AI and
[1] 6 Indonesian high Semi-structured It highlights the LLMs in order to
school students interview potential of ChatGPT as harness their
a learning tool to capabilities effectively
promote autonomy and for educational
personalized learning purposes.
among English [8] ChatGPT – This article highlights
Language Learners the potential of
(ELLs) in the ChatGPT as a
Kurikulum Merdeka personalized language
Belajar (KMB) learning partner that
program. The findings provides feedback and
suggest that ChatGPT practice. However,
can enhance students’ there is a concern that
ability to take charge of ChatGPT may lead to
their learning limited or zero learning
processes, set if students rely too
objectives, and make heavily on it. It
decisions about emphasizes the need
language acquisition. for finding a balance
[2] 15 4th grade students Online Findings revealed that between assistance and
studying in a public questionnaire students found independent learning.
school in Türkiye in ChatGPT to be [10] 44 undergraduate Pretest-posttest The integration of
the 2022–2023 engaging, enjoyable, students (20 males quasi-experimental argumentative
academic year and beneficial for and 24 females) from design chatbots into classroom
academic achievement. a comprehensive debates can enhance
It provided accurate university in China argumentation skills
and prompt responses and increase task
to student queries, motivation among
offering a greater students participated in
amount of information the study.
compared to [11] ChatGPT – ChatGPT presents
traditional resources. significant
The students opportunities for
recommended its use in improving second/
other subjects such as foreign language
social studies, teaching and
mathematics, and assessments. It offers a
Turkish. personalized learning
[3] 80 primary school Questionnaire The findings showed experience and opens
students and that ChatGPT was able up avenues for
instructors of the to motivate learners to research. Educators
English language develop reading and should engage in
writing skills, while discussions with
attitudes towards its students about the
effect on listening and ethical and responsible
speaking skills were application of chatbots
neutral. The study rather than avoiding
suggests that ChatGPT- the topic altogether.
based teaching is [12] ChatGPT Poetry analysis While AI models like
motivational and ChatGPT-3 can imitate
should be used as a the style and language
learning tool, with of famous poets, they
further research are unable to generate
needed to explore its emotions they have not
advantages and address experienced. The
any potential negative article suggests
effects. reframing discussions
[5] ChatGPT – LLMs, such as about AI’s impact on
ChatGPT, can assist humanity with a focus
teachers by reducing on understanding the
their preparation time uniquely human
and providing anxiety and the drive to
personalized feedback create.
to students. The article [13] ChatGPT – The authors suggest
highlights the various that GenAI applications
tasks that LLMs can can enhance
perform, such as personalized learning,
summarizing texts, promote creativity,
correcting grammar, critical thinking, and
generating writing problem-solving skills.
prompts, and They also emphasize
generating lesson the importance of
ideas. It emphasizes the developing guidelines
importance of and templates for
GenAI-based learning
(continued on next page)

6
L. Law Computers and Education Open 6 (2024) 100174

Table 3 (continued ) Table 3 (continued )

Author Data/Participants Instrument Summary Author Data/Participants Instrument Summary

and conducting high- assessment, are also

quality studies to discussed.
further explore its [19] ChatGPT; 6 male and Questionnaire; The study emphasizes
effectiveness and 6 female University semi-structured that instructors should
potential in instructors interviews possess digital
educational settings. competencies and
[14] 11 English language Semi-structured The study identifies pedagogical knowledge
teachers from ten interviews and the four roles of ChatGPT to effectively
elementary schools in teachers’ interaction (interlocutor, content implement AI-driven
South Korea logs with the provider, teaching teaching tools. It also
chatbot assistant, and stresses the importance
evaluator) and three of providing tailored
roles of teachers support and
(orchestrating professional
resources, making development programs
students active to address the
investigators, and challenges and
raising AI ethical concerns associated
awareness). It with adopting AI in the
emphasizes the classroom.
importance of teachers’ [20] ChatGPT – It emphasizes the
pedagogical knowledge benefits of utilizing
when using AI tools ChatGPT in material
and provides development and
implications for the assessment, including
future use of LLM- the generation of
powered chatbots in tailored text passages
education. and comprehension
[15] 32 empirical studies Systematic review The review suggests the questions. However, it
on speech-recognition need for further also acknowledges the
chatbots for language research on speech- need for further
learning were recognition chatbots, empirical research to
reviewed particularly regarding evaluate the
the use of LLMs. effectiveness and
[16] Various AI software: – The article discusses quality of ChatGPT in
Intelligent and the evolution of text language education.
Interactive Writing revision tools, from [23] 121 Grade five and 6 A pre-post design, The researchers
Assistant rule-based approaches students (aged 11–12) including both the examined the impact of
(In2Writing2) to deep neural-based in elementary schools experimental and using an artificial
ones, and located in Seoul, comparison intelligence-based
acknowledges the Gwangju and conditions content generator
existing challenges in Jeollanam-do, South (AICG) on young
terms of accessibility, Korea English-as-a-foreign-
context consideration, language learners’
and discursive reading enjoyment and
information. interest. The findings
[17] 109 respondents Questionnaire The study highlights suggest that AICG
across the Czech the importance of technology has the
Republic simplicity, quick potential to improve
practice opportunities, second-language
gamification features, learning experiences.
accessibility, and cost- [24] 45 YouTube videos This study identified
free usage in chatbots that ChatGPT is a
for language learning. valuable tool for
It also emphasizes the language teaching but
significance of cannot replace teachers
immediate feedback, completely. The study
user-friendly also highlighted two
interfaces, and short- gaps: the learning
term interactions. optimization gap and
[18] ChatGPT – It highlights the the knowledge
benefits of ChatGPT in comprehension gap.
providing authentic [25] GPT-3 model (text- The study found that
interactions, davinci-003) GPT-3 performed well
explaining word in GEC tasks,
meanings, generating outperforming existing
texts in various genres, supervised and
and offering unsupervised
vocabulary support. approaches.
However, debates and Additionally, the
drawbacks surrounding controllability of GPT-3
the ethical use of in terms of minimal
ChatGPT, including edits, fluency edits, and
concerns about learner levels was
cheating and demonstrated.
(continued on next page)

7
L. Law Computers and Education Open 6 (2024) 100174

Table 3 (continued ) Table 3 (continued )

Author Data/Participants Instrument Summary Author Data/Participants Instrument Summary

[27] 4 writing teachers Semi-structured The article states that [35] – – The article suggests
from three different interviews integrating AI writing that whether the use of
universities; Quillbot, tools in EFL classrooms these tools constitutes
WordTune, Jenni, positively impacts plagiarism or a breach
Chat-GPT, Paperpal, students’ writing of academic integrity
Copy.ai, and Essay quality, especially in depends on how
Writer terms of content and transparent students
organization. are about their usage. It
[28] 12,100 essays GPT-3 text-davinci- The results indicated highlights the need for
contained in the ETS 003 model that AES using GPT-3 Higher Education
Corpus of Non-Native demonstrated a certain Institutions to update
Written English level of accuracy and their academic
(TOEFL11) reliability, with the integrity policies to
potential to enhance address the use of LLMs
scoring accuracy by in educational
incorporating linguistic environments.
features, suggesting [36] ChatGPT – The article emphasizes
that AI language the need for a
models such as GPT-3 multidisciplinary and
can drastically change collaborative approach
writing evaluation and to address the ethical,
feedback in both societal, and practical
research and practice. challenges associated
[29] 10 English as a Interviews The faculty members with artificial
Foreign Language had varying opinions intelligence (AI). It
(EFL) faculty about ChatGPT, with highlights the
members at Northern some acknowledging importance of robust AI
Border University its usefulness in policies, improved AI
providing immediate literacy, sustainable AI
responses to questions, practices, and ongoing
while others worried dialog.
about its impact on [38] – – It highlights the need
critical thinking and for comprehensive
research skills. pedagogical
[30] 67 university teachers Questionnaire & Teachers hold diverse approaches and
(Questionnaire), 23 interview opinions about the academic integrity
university teachers advantages of AI policies that address
(Interview) technologies for the use of AI-powered
students, expressing writing tools beyond
both positive and large language models
negative views on its (LLMs) like ChatGPT.
impact on academic The study categorizes
integrity. The study three types of digital
discusses the writing tools: machine
importance of translators (MTs),
providing digital writing
comprehensive assistants (DWAs), and
training and support to automated
teachers, enabling paraphrasing tools
them to effectively (APTs).
incorporate AI while [39] ChatGPT Case studies, The article suggests
upholding academic multiple-choice that while ChatGPT has
honesty. Additionally, tests, writing tasks, great potential in
ethical considerations and mathematics enhancing language
and guidelines are queries for acquisition, users
urged to ensure the standardized tests should exercise critical
responsible discernment, and
implementation of AI educators need to
in education. continually update
[31] 350 students and Survey, ANOVA, The results indicated a their understanding of
teachers post hoc multiple significant correlation information
comparison tests between teachers’ and technology for effective
students’ perceptions implementation.
of using ChatGPT, and [40] 69 Japanese Questionnaire; The article suggests
it was found that university students chatgpt, that ChatGPT is the
ChatGPT had Grammarly, most helpful grammar-
substantial impact on prowritingaid checking tool for
student motivation and Japanese English
engagement. The Language Learners
research suggests that (ELLs) compared to
incorporating ChatGPT Grammarly and
into the educational ProWritingAid. The
system can improve article discusses the
student learning potential of ChatGPT in
outcomes. improving students’
(continued on next page)

8
L. Law Computers and Education Open 6 (2024) 100174

Table 3 (continued ) Table 3 (continued )

Author Data/Participants Instrument Summary Author Data/Participants Instrument Summary

understanding and use [50] 67 EFL students from Open-ended This article suggests
of English grammar. four Hong Kong questions; thematic that implementing
[42] ChatGPT – The article emphasizes secondary schools analysis machine-in-the-loop
that while AI tools like writing in EFL
ChatGPT, Bing AI, and classrooms has both
DALL-E have the benefits and
potential to benefit challenges. The study
gifted education highlights the
programs by providing importance of aligning
personalized learning, activity goals with
advanced content, and students’ values,
creative opportunities, language abilities, and
teachers need to AI capabilities to
carefully consider the enhance students’
advantages and activity systems.
disadvantages before [51] 23 Hong Kong Multiple linear This article suggests
incorporating them secondary school regression, cluster that the use of AI-
into the classroom. students analyses, syntactic generated text can
Despite the limitations complexity of the improve the quality of
of AI, it can be a stories’ AI- both high-scoring and
valuable tool for generated text low-scoring students’
meeting the needs of writing, providing
gifted and talented insights for
students. pedagogical strategies
[44] ChatGPT – The article proposes a in using AI-generated
framework for text for EFL students’
developing a foreign writing.
language teaching [52] 16 Chinese Classroom In terms of L2 writing
software tool for undergraduate observations, pedagogy, the study’s
children using majoring in EFL learning log findings indicated that
Augmented Reality analysis, and ChatGPT has the
(AR), Voicebots, and interviews potential to bring
ChatGPT. It suggests benefits, such as
that the framework and enhancing writing
design principles efficiency. However,
presented can serve as participants raised
a blueprint for highly concerns regarding
effective foreign academic integrity and
language teaching fairness in education.
software. The study emphasized
[46] 58 senior high schools Questionnaire; in- This article found that the importance of
students (25 males depth interviews via while AI-based reassessing plagiarism
and 33 females) in mobile instant learning tools have in the age of AI and
Semarang, Indonesia messaging shown potential in establishing guidelines
assisting students with and policies to ensure
academic research and appropriate use of the
drafting, they have not tool.
significantly improved [54] Google Scholar, – The findings revealed
the overall quality of ProQuest, IEEE, gaps in the design of AI
students’ academic ScienceDirect, and dialog systems, such as
papers. The study Web of Science. the neglect of debate
recommends and problem-solving
enhancing AI-based skills and the absence
tools by adding of cultural, humorous,
features for editing and empathetic
Indonesian text and functions. The study
improving AI literacy. suggests focusing on
[48] 12 faculty members (9 – Positive perceptions meaning-based
females and 3 males) towards the use of AI communication,
from a private were observed among intelligibility in
university in Chile the teachers, who language competency,
acknowledged its and problem-solving
potential for improving skills in future
learning and teaching. research.
The study highlights [55] 30 Chinese English Semi-structured This study found that
the significance of students (25 females interviews, pre-post although the chatbot
taking into account and 5 males) at a tests of fallacy was considered to be
students’ motivation university in Hong knowledge, and pre- somewhat less effective
levels and teachers’ Kong post motivation in enhancing target
technological and questionnaires knowledge, it
pedagogical demonstrated greater
competence when effectiveness in
incorporating AI into enhancing learner
the EFL classroom. motivation. The study
emphasized the
benefits of chatbots in
(continued on next page)

9
L. Law Computers and Education Open 6 (2024) 100174

Table 3 (continued ) 4. Discussion

Author Data/Participants Instrument Summary
4.1. Summary of evidence
terms of their
interactions between
humans and From an initial 224 entries from the selected databases, 41 studies
computers, the ability have met the inclusion criteria and were investigated in this scoping
to create study plans, review. The objective is to determine the current state of research on the
and their accessibility.
use of GenAI in language teaching and learning, and to identify research
[56] 40 sophomores Coh-Metrix: A tool This study suggests that
majoring in English for analysing while ChatGPT gaps and areas for future investigation. Four review questions were
(10 males and 30 language discourses, outperformed human conceptualised to meet the objective and were successfully answered.
females) (native used to measure the writers in certain As far as key terms surrounding the use of GenAI in language edu­
Chinese speakers) in a data in terms of five aspects of writing, such cation is concerned (see Section 3.3.1 Review question 1), there is a
top-20 university in discourse as narrativity, word
China components. concreteness, and
variety of synonyms and acronyms used by the authors. To a certain
referential cohesion, it extent, this can lower the chances of their papers appearing on online
fell behind Chinese databases and may even cause confusion to readers. Therefore, there is a
intermediate English need for a more unified term to maximize the exposure of studies, and a
(CIE) learners in terms
simpler term such as generative artificial intelligence (GenAI) programs
of syntactic simplicity
and deep cohesion. may be more practical.
[57] ChatGPT SWOT analysis The authors put Studies included in this current review have demonstrated the
forward the idea of versatility of GenAI tools at different education levels and languages (see
conducting a SWOT Section 3.3.2 Review question 2). However, many papers have
analysis on ChatGPT
and provide
emphasised on EFL and in university settings. To broaden the reader
recommendations on base, it can be beneficial to add more research on GenAI for teaching
effectively other purposes of English or other languages at different education
incorporating it into levels.
educational settings.
Given the time of publication of the selected papers (i.e., almost all in
The research
underscores the the first half of 2023), it is not surprising that researchers have focused
importance of on T&L and related policies, writing, and assessments (see Section 3.3.3
acknowledging Review question 3a). Indeed, these are the areas which are impacted by
ChatGPT’s limitations the launch of the freely available OpenAI’s ChatGPT (GPT-3), a then-
while capitalizing on
text-based chatbot [7]. Since multimodal GenAI programs are now
its strengths to improve
the field of education. available [e.g., GPT-4 [33], DALL⋅E 2 [34]], it is expected that the areas
[58] 30 EFL teachers and Questionnaire The findings of research and disciplines to be significantly broader than it currently is
431 students from the surveys, semi- demonstrated that the within a very short period of time. Generally, researchers are positive
English Department structured implementation of AI-
towards the use of GenAI in language education (see Section 3.3.3 Re­
and the Beninese interviews, powered collaborative
Center for Foreign observation of and interactive view question 3b), but educators should be aware of the environmental
Languages of the online language language learning has impacts of GenAI, and that some potential effects of this technology race
University of learning sessions the ability to enhance in GenAI development is yet to surface (see [22,41]).
Abomey-Calavi EFL teaching in the This review encompasses a range of studies examining the potential
post-pandemic online
advantages and challenges associated with the implementation of GenAI
setting. It was shown
that incorporating AI- in language teaching and learning (see Section 3.3.4 Review question 4).
assisted collaborative These impacts can be classified into two main categories: psychological
online learning leads to aspects and productivity considerations. From a psychological
improved engagement
perspective, several studies have indicated that students may derive
in EFL instruction,
enhanced learning
various benefits from utilizing GenAI programs. These programs can
outcomes for EFL adapt to learners’ proficiency levels, learning pace, and cognitive abil­
students, and higher ities, and thus may function as virtual private tutors, offering tailored
levels of teacher language learning suggestions and personalized feedback on-demand (e.
satisfaction.
g., [57]), thereby enhancing students’ motivation, creativity, interest,
learning experience, and autonomy (e.g., [48,55,58]). These advantages
writing that is undetectable by existing plagiarism detection software), surpass the capabilities of human class teachers. However, more
and learner motivation impairment (i.e., ChatGPT produces writing research is needed to investigate their effectiveness in different educa­
without students putting in significant effort). Studies conducted on tional settings. From a perspective of productivity gains, GenAI pro­
university teachers’ perception of GenAI programs reveal teachers’ grams can swiftly generate drafts, compose short essays, correct
concern about academic integrity [29,30], reinforcement of biases or grammatical errors, and refine sentences and paragraphs, effectively
misinformation, as well as hindrance of students’ development of crit­ saving valuable time. Consequently, teachers also stand to benefit from
ical thinking and research skills [29]. GenAI programs. In terms of productivity, these programs can facilitate
Another important aspect of ethical concern is data privacy and se­ idea generation, provide examples, aid in grading, and offer feedback to
curity. However, this topic has only been discussed at length by [29] and students. Psychologically, the increased productivity resulting from
[31] and [57]. The major argument is that as AI-powered chatbots GenAI utilization may alleviate stress among teachers and afford them
collect and analyze user data to provide personalized learning experi­ additional time and resources to enhance the quality of their instruction.
ences, there is a need to ensure that students’ personal information is However, it is worth noting that contrasting viewpoints have
protected and handled in a secure manner. emerged from certain papers, positing that the same psychological as­
pects (once again referring to students’ motivation, creativity, interest,
learning experience, and autonomy) may potentially be compromised
by the integration of GenAI programs. These papers argue that these

10
L. Law Computers and Education Open 6 (2024) 100174

programs may be perceived as shortcuts for writing [30,52], and con­ their potential benefits in terms of enhancing productivity and psycho­
cerns have been raised about the potential lack of contextual under­ logical aspects may be of relevance to other disciplines that incorporate
standing and the inability to assess higher-order cognitive skills in GenAI technologies. The findings highlight the broader implications of
generative AI systems [29], thereby increasing the risk of plagiarism and GenAI tools in education and emphasize the need for a comprehensive
contributing to academic dishonesty [52]. Interestingly, however, that understanding of their impact on learners and educators.
when examining the issue through the lens of productivity, the potential The ethical considerations raised in this scoping review, particularly
advantages offered by GenAI programs remain uncontested. regarding data privacy and security, also hold significance for policy­
Although ethical concerns (e.g., algorithmic bias, plagiarism, makers and educators alike. The limited attention given to this aspect in
dishonesty, inequality) have been discussed by a few researchers in their the reviewed literature underlines the importance of raising awareness
papers, there is a clear lack of attention given to an important aspect of and ensuring adequate teacher ‘GenAI literacy’ to address potential risks
ethical consideration: data privacy and security. Only six papers [19,20, and safeguard student data. Policymakers can utilize these findings to
29,31,48,57] briefly mention this topic within the selected literature. develop guidelines and policies that promote ethical practices and
Among them, [57] provide the only substantial discussion, addressing protect the privacy of individuals involved in GenAI-based language
potential risk factors such as security breaches or hacking that could lead education.
to illegal access to student data. They also raise concerns about the
misuse of student data by third parties and emphasize the role of 4.3. Limitations
teachers in safeguarding the data and obtaining informed consent from
students before using GenAI programs. One apparent limitation is the year span of publication of this study.
In addition, the lack of AI literacy discussion in the selected papers While publications spanning across 6 years can be considered a
could potentially suggest the lack of awareness among language re­ reasonable length (i.e., they must be published between 2017 and 2023,
searchers/educators. One highly possible scenario, but has not been up till the time of this writing on July 25, 2023), studies on GenAI are
discussed in the selected literature, is that teachers may unknowingly expected to surge after the cut-off date. This means many potentially
introduce seemingly beneficial GenAI programs to students without useful articles have not been considered in the current study. Consid­
fully understanding how the first-party data, rather than third-party ering the rapid development of GenAI, I posit that it is worth performing
data, is collected and utilized. As discussed in a recent publication by a similar literature review to this current study on a yearly basis in order
[7], it is crucial that all educators should have a satisfactory level of to keep up with the latest trend in the related scholarship.
teacher AI literacy. Another important limitation to note is that the literature lacks
Logically then, a teacher’s knowledge in data privacy and security empirical evidence on the effectiveness and quality of students’ work
within ‘GenAI literacy’ (which is emphasised here not as a different type after incorporating GenAI tools. This means that there may still be un­
of AI literacy, but a subset of it) is of utmost importance, since a teacher expected issues or uncertainties regarding the pedagogical impact of
should be able to educate students that any user data – be it text, image, GenAI in educational contexts.
sound, or video – once uploaded, has the potential to be used by the
programs for training and replication (see [26] for a video with 5. Conclusion
AI-generated voice-over of US Presidents playing a Nintendo game, and
see [37] for a demonstration video of how GenAI generates voices). In The application of GenAI in language teaching and learning is a
short, every teacher should serve as the first line of defense to ensure promising area of research with the potential to transform language
that students understand the concept of ‘what they pay’ before ‘what education. This scoping literature review has achieved its objective by
they get’. answering several review questions with respect to the literature,
revealing key terms related to GenAI in language education, most
4.2. Significance of findings researched language studies and education levels, areas of research,
attitude towards the use of GenAI, and potential benefits and challenges
This scoping review provides insights into the research types and of GenAI application in the context.
methods used, the positive attitudes towards GenAI tools in language The results of this scoping review highlight several implications and
teaching and learning, the focus on various language skills, the assess­ future directions for research and practice. Firstly, there is a clear need
ment and material development aspects, and the policies related to their for more empirical studies to provide a comprehensive understanding of
implementation. These findings help shed light on the current state of the short and long-term effectiveness and impact of GenAI tools. This
research, identify gaps, and offer implications for future investigations. includes exploring both text-based and multimodal-based tools and their
The findings have relevance to key groups involved in language specific applications in language education. Secondly, continuous and
education. Language learners and teachers can take into account the regular investigations are required to explore the ethical considerations
positive attitudes reported towards the use of GenAI programs such as and potential limitations of these fast-changing technologies. As GenAI
ChatGPT and other GenAI tools and consider incorporating these pro­ tools continue to advance, there is a need to address concerns related to
grams into their lessons. The identification of language skills and areas data privacy, security, and the responsible use of these technologies in
of language learning that have been investigated may guide curriculum educational settings. Thirdly, future research should focus on specific
development and instructional practices. Additionally, developers of language skills, such as writing or speaking in different languages, to
GenAI technologies may use the review findings to inform the design provide targeted interventions using GenAI tools. By understanding how
and improvement of GenAI tools for language teaching and learning. these tools can support and enhance specific language skills, educators
Furthermore, the findings of this scoping review hold significance for can develop more effective instructional strategies. Lastly, stakeholder
researchers in the field of language education. The identified research engagement is crucial in shaping the implementation and use of GenAI
gaps, such as the need for more studies on GenAI for purposes other than programs [19]. Students, language educators, researchers, policy­
English language teaching, provide valuable directions for future in­ makers, and developers should collaborate to ensure that these tools are
vestigations. Researchers may also draw insights from the various integrated into language teaching and learning contexts in a meaningful
research types and methods employed in the reviewed studies, allowing and responsible manner. Moreover, the implications of GenAI programs
for the advancement of research methodologies in the domain of GenAI extend beyond language education, making it important to consider
in language teaching and learning. their integration in other subjects as well. Since GenAI is here to stay,
Moreover, the implications of this scoping review extend beyond educators should consider incorporating GenAI tools into their teaching
language education. The positive attitudes towards GenAI programs and practices. However, it is important to remain vigilant and mindful of the

11
L. Law Computers and Education Open 6 (2024) 100174

potential risks involved. Continuous professional development for edu­ [11] Hong WCH. The impact of ChatGPT on foreign language teaching and learning:
opportunities in education and research. J Educ Technol Innov 2023;5(1):37–45.
cators on GenAI-related subjects is paramount to ensure that they are
https://jeti.thewsu.org/index.php/cieti/article/view/103/64.
equipped with the necessary knowledge and skills to make informed [12] Hutson J, Schnellmann A. The poetry of prompts: the collaborative role of
decisions and effectively integrate these tools into their teaching prac­ generative artificial intelligence in the creation of poetry and the anxiety of
tices [7]. machine influence. Glob J Comput Sci Technol 2023;23(1). https://digitalcommo
ns.lindenwood.edu/cgi/viewcontent.cgi?article=1462&context=faculty-research
In conclusion, the findings from this scoping review contribute to the -papers.
existing knowledge on the use of GenAI tools in language education. The [13] Hwang GJ, Chen NS. Editorial position paper: exploring the potential of generative
identified concepts, themes, and types of evidence provide an overview artificial intelligence in education: applications, challenges, and future research
directions. Educ Technol Soc 2023;26(2). https://doi.org/10.30191/ETS.202304_
of the current research landscape, and have implications for major 26(2).0014.
stakeholders involved in language education. Future research should [14] Jeon J, Lee S. Large language models in education: a focus on the complementary
build upon these findings to further explore the potential of GenAI tools relationship between human teachers and ChatGPT. Educ Inf Technol (Dordr)
2023. https://doi.org/10.1007/s10639-023-11834-1.
in enhancing language teaching and learning experiences. [15] Jeon J, Lee S, Choi S. A systematic review of research on speech-recognition
chatbots for language learning: implications for future directions in the era of large
language models. Interactive Learn Environ 2023. https://doi.org/10.1080/
Declaration of Generative Ai and AI-assisted technologies in the 10494820.2023.2204343.
writing process [16] Jourdan, L., Boudin, F., Dufour, R., & Hernandez, N. (2023). Text revision in
scientific writing assistance: an overview. arxiv.org.
[17] Klimova, B., Pikhart, M., & Al-Obaydi, L.H. (2023). The use of persona in foreign
During the preparation of this work the author used ChatGPT in
language learning facilitated by chatbots. researchsquare.com https://www.res
order to improve readability and language. After using this tool/service, earchsquare.com/article/rs-3129096/latest.pdf.
the author reviewed and edited the content as needed and take full re­ [18] Kohnke L, Moorhouse B, Zou D. ChatGPT for language teaching and learning. RELC
J 2023;54(2):1–14. https://doi.org/10.1177/00336882231162868.
sponsibility for the content of the publication.
[19] Kohnke L, Moorhouse B, Zou D. Exploring generative artificial intelligence
preparedness among university language instructors: a case study. Comput Educ
Artif Intell 2023;5:100156. https://doi.org/10.1016/j.caeai.2023.100156.
CRediT authorship contribution statement
[20] Koraishi O. Teaching English in the age of AI: embracing ChatGPT to optimize EFL
materials and assessment. Lang Educ Technol (LET J) 2023;3(1):55–72. https://l
Locky Law: Writing – review & editing, Writing – original draft, angedutech.com/letjournal/index.php/let/article/download/48/37.
Visualization, Validation, Software, Resources, Project administration, [21] Laupichler MC, Aster A, Schirch J, Raupach T. Artificial intelligence literacy in
higher and adult education: a scoping. Comput Educ Artif Intell 2022;3:100101.
Methodology, Investigation, Formal analysis, Data curation, https://doi.org/10.1016/j.caeai.2022.100101.
Conceptualization. [22] Law L. Impact and implications of AI in education for every (language) teacher
[Review of the journal article the future of AI and education: some cautionary, by
Selwyn, N.]. STILE - Scholarsh Teach Lang Educ 2023;1(1):9. https://www.stile.
hk/osp/article/view/77.
Declaration of competing interest [23] Lee J, Shin D, Noh W. Artificial intelligence-based content generator technology for
young English-as-a-foreign-language learners’ reading enjoyment. RELC J 2023;54
(2). https://doi.org/10.1177/00336882231165060.
The authors declare that they have no known competing financial [24] Li B, Bonk C, Kou X. Exploring the multilingual applications of ChatGPT:
interests or personal relationships that could have appeared to influence uncovering language learning affordances in youtuber videos. Int J Comput Assist
the work reported in this paper. Lang Learn Teach 2023;13(1). https://doi.org/10.4018/IJCALLT.326135.
[25] Loem, M., Kaneko, M., Takase, S., & Okazaki, N. (2023). Exploring effectiveness of
GPT-3 in grammatical error correction: a study on performance and controllability
Acknowledgements in prompt-based methods. arXiv https://arxiv.org/pdf/2305.18156.
[26] Mantekilla. (2023, March 12). Trump and the boys play Splatoon 3 (Full) [Video].
YouTube. https://youtu.be/sCcpUgj9TKo.
I would like to thank the guest editors of this special issue and my [27] Marzuki, Widiati U, Rusdin D, Indrawati D, Indrawati I. The impact of AI writing
reviewers for their valuable comments and feedback. tools on the content and organization of students’ writing: EFL teachers’
perspective. Cogent Educ 2023;10(2). https://doi.org/10.1080/
2331186X.2023.2236469.
References [28] Mizumoto A, Eguchi M. Exploring the potential of using an AI language model for
automated essay scoring. Res Methods Appl Linguist 2023;2(2):100050. https://
doi.org/10.1016/j.rmal.2023.100050.
[1] Agustini NPO. Examining the role of ChatGPT as a learning tool in promoting
[29] Mohamed A. Exploring the potential of an AI-based Chatbot (ChatGPT) in
students’ English language learning autonomy relevant to Kurikulum Merdeka
enhancing English as a foreign language (EFL) teaching: perceptions of EFL faculty
Belajar. EDUKASIA: J Pendidikan Dan Pembelajaran 2023;4(2):921–34. https://jur
members. Educ Inf Technol (Dordr) 2023. https://doi.org/10.1007/s10639-023-
naledukasia.org/index.php/edukasia/article/download/373/281.
11917-z.
[2] Aktay S, Seçkin G, Uzunoğlu D. ChatGPT in Education. Türk Akademik Yayınlar
[30] Mohammadkarimi E. Teachers’ reflections on academic dishonesty in EFL
Dergisi (TAY J) 2023;7(2):378–406. https://doi.org/10.29329/
students’ writings in the era of artificial intelligence. J Applied Learn Teach 2023;6
tayjournal.2023.543.03.
(2). https://doi.org/10.37074/jalt.2023.6.2.10.
[3] Ali J, Shamsan M, Hezam T, Mohammed A. Impact of ChatGPT on learning
[31] Muñoz S, Gayoso G, Huambo A, Tapia R, Incaluque J, Aguila O, Cajamarca J,
motivation: teachers and students’ voices. J Engl Stud Arabia Felix 2023;2(1):41–9.
Acevedo J, Rivera H, Arias-Gonzáles J. Examining the Impacts of ChatGPT on
https://doi.org/10.56540/jesaf.v2i1.51.
student motivation and engagement. Soc Space J 2023;23(1). https://socialspacejo
[4] Anthony L. AntConc (Version 4.2.2) [Computer software]. Tokyo, Japan: Waseda
urnal.eu/menu-script/index.php/ssj/article/download/156/68.
University. Available from; 2023. https://www.laurenceanthony.net/software.
[32] OpenAI. New GPT-3 capabilities: edit & insert. March 15. OpenAI; 2022. Retrieved
[5] Bonner E, Lege R, Frazier E. Large language model-based artificial intelligence in
December 14, 2023, from, https://openai.com/blog/gpt-3-edit-insert.
the language classroom: practical ideas for teaching. Teach Engl Technol 2023;23
[33] OpenAI. GPT-4. March 14. OpenAI; 2023. Retrieved December 10, 2024, from, http
(1):23–41. https://doi.org/10.56297/BKAM1691/WIEO1749.
s://openai.com/research/gpt-4.
[6] Chen X, Zou D, Xie H, Cheng G, Liu C. Two Decades of Artificial Intelligence in
[34] OpenAI. (n.d.). DALL⋅E 2. OpenAI. Retrieved December 10, 2023, from https://
Education. Educ Technol Soc 2022;25(1):28–47. https://www.jstor.org/stable/
openai.com/dall-e-2.
48647028.
[35] Perkins M. Academic integrity considerations of AI large language models in the
[7] Chiu TKF. The impact of Generative AI (GenAI) on practices, policies and research
post-pandemic era: ChatGPT and beyond. J Univ Teach Learn Pract 2023;20(2):07.
direction in education: a case of ChatGPT and Midjourney. Interact Learn Environ
https://doi.org/10.53761/1.20.02.07.
2023. https://doi.org/10.1080/10494820.2023.2253861.
[36] Ray PP, Das PK. Charting the terrain of artificial intelligence: a multidimensional
[8] Forrester A. ChatGPT: what does it mean for language centres?. STiLE-Scholarship
exploration of ethics, agency, and future directions. Philos Technol 2023;36:40.
of teaching in language education. 2023. https://www.stile.hk/pdf/Chatgpt%
https://doi.org/10.1007/s13347-023-00643-6.
20what%20does%20it%20mean%20for%20language%20centres.pdf.
[37] Resemble A.I. (2022, December 13). How to use Generative AI (GPT-3 and AI Voices)
[9] Google. What is generative ai and what are its applications? Google Cloud; 2023.
in 15 Min. [Video]. YouTube. https://youtu.be/tnNcaFcLkh8.
Retrieved February 25, 2024, from, https://cloud.google.com/use-cases/gene
[38] Roe J, Renandya WA, Jacobs GM. A review of AI-Powered writing tools and their
rative-ai.
implications for academic integrity in the language classroom. J Engl Appl Linguist
[10] Guo K, Zhong Y, Li D, Chu S. Effects of chatbot-assisted in-class debates on
2023;2(1):3. https://doi.org/10.59588/2961-3094.1035.
students’ argumentation skills and task motivation. Comput Educ 2023;203:
104862. https://doi.org/10.1016/j.compedu.2023.104862.

12
L. Law Computers and Education Open 6 (2024) 100174

[39] Sakai N. Investigating the feasibility of ChatGPT for personalized English language [50] Woo D, Susanto H, Guo K. EFL students’ attitudes and contradictions in a machine-
learning: a case study on its applicability to Japanese students. osf.io 2023. https in-the-loop activity system. ResearchGate 2023. https://www.researchgate.net/
://osf.io/cv9f2/download. profile/David-Woo-3/publication/372315207_EFL_Students’_Attitudes_and_Cont
[40] Schmidt-Fajlik R. ChatGPT as a grammar checker for Japanese English language radictions_in_a_Machine-in-the-loop_Activity_System/links/64af930f95bbbe0c6e
learners: a comparison with Grammarly and ProWritingAid. AsiaCALL Online J 2f84f7/EFL-Students-Attitudes-and-Contradictions-in-a-Machine-in-the-loop-Ac.
2023;14(1):105–19. https://doi.org/10.54855/acoj.231417. [51] Woo, D., Susanto, H., Yeung, C., Guo, K., & Fung, A. (2023b). Exploring AI-
[41] Selwyn N. The future of AI and education: some cautionary notes. Eur J Educ 2022; Generated text in student writing: How does AI help? arXiv https://arxiv.org/pdf/2
57(4):620–31. https://doi.org/10.1111/ejed.12532. 304.02478.
[42] Siegle D. A Role for ChatGPT and AI in gifted education. Gifted Child Today 2023; [52] Yan D. Impact of ChatGPT on learners in a L2 writing practicum: an exploratory
46(3). https://doi.org/10.1177/10762175231168443. investigation. Educ Inf Technol (Dordr) 2023. https://doi.org/10.1007/s10639-
[43] Sparrow B, Liu J, Wegner DM. Google effects on memory: cognitive consequences 023-11742-4.
of having information at our fingertips. Science (1979) 2011;333(776):776–8. [53] Zawacki-Richter O, Marín VI, Bond M, Gouverneur F. Systematic review of
https://doi.org/10.1126/science.1207745. research on artificial intelligence applications in higher education – where are the
[44] Topsakal O, Topsakal E. Framework for a foreign language teaching software for educators? Int J Educ Technol High Educ 2019;16:39. https://doi.org/10.1186/
children utilizing AR, Voicebots and ChatGPT (Large Language Models). J Cogn s41239-019-0171-0.
Syst 2022;7(2):33–8. https://dergipark.org.tr/en/download/article-file/2864638. [54] Zhai C, Wibowo S. A systematic review on artificial intelligence dialogue systems
[45] Tricco AC, Lillie E, Zarin W, O’Brien KK, Colquhoun H, Levac D, Moher D, for enhancing English as foreign language students’ interactional competence in
Peters MDJ, Horsley T, Weeks L, Hempel S, Akl EA, Chang C, McGowan J, the university. Comput Educ Artif Intell 2023;4:100134. https://doi.org/10.1016/
Stewart L, Hartling L, Aldcroft A, Wilson MG, Garritty C, Straus SE. PRISMA j.caeai.2023.100134.
extension for scoping reviews (PRISMA-SCR): checklist and explanation. Ann [55] Zhang R, Zou D, Cheng G. Chatbot-based learning of logical fallacies in EFL
Intern Med 2018;169(7):467–73. https://doi.org/10.7326/M18-0850. writing: perceived effectiveness in improving target knowledge and learner
[46] Utami SPT, Andayani, Winarni R, Sumarwati. Utilization of artificial intelligence motivation. Interact Learn Environ 2023. https://doi.org/10.1080/
technology in an academic writing class: how do Indonesian students perceive? 10494820.2023.2220374.
Contemp Educ Technol 2023;15(4):ep450. https://doi.org/10.30935/cedtech/ [56] Zhou, T., Cao, S., Zhou, S., Zhang, Y., & He, A. (2023). Chinese intermediate
13419. English learners outdid ChatGPT in deep cohesion: evidence. arXiv https://arxiv.
[47] Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, Kaiser Ł, org/pdf/2303.11812.
Polosukhin I. In: 31st Conference on Neural Information Processing Systems (NIPS [57] Zhu C, Sun M, Luo J, Li T, Wang M. How to harness the potential of ChatGPT in
2017); 2017. https://proceedings.neurips.cc/paper/2017/file/3f5ee243547dee91 education? Knowl Manag E-Learn 2023;15(2):133–52. https://doi.org/10.34105/j.
fbd053c1c4a845aa-Paper.pdf. kmel.2023.15.008. https://www.researchgate.net/profile/Minhong-Wang-2/
[48] Vera F. Integrating artificial intelligence (AI) in the EFL classroom: benefits and publication/370894024_and_Technology/links/646789029533894cac7e6771/an
challenges. Transf Electron J 2023;4(2). https://revistatransformar.cl/index.php/t d-Technology.pdf.
ransformar/article/download/93/49. [58] Zounhin Toboula, C.M. (2023). Exploring the impact of AI-Powered collaborative
[49] Wharton School. Practical ai for instructors and students part 1: introduction to ai for and interactive NLP apps on EFL teaching in the post-Covid-19 era https://papers.
teachers and students [Video]. July 31. YouTube; 2023. https://www.youtube. ssrn.com/sol3/papers.cfm?abstract_id=4398817.
com/watch?v=t9gmyvf7JYo&list=PLwRdpYzPkkn302_rL5RrXvQE8j0jLP
02j&index=1.

13