INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 55
www.rsisinternational.org
The Use of ChatGPT AI Audio in Mastering Arabic Speaking Skills
at the Beginner Level
Wan Mohd Hamdi Bin Wan Nordin@Wan Ghazali, Adhanawati Abd Rapa
*
, Ijlal Saja
Academy of Language Studies, University of MARA Technology, Melaka Branch, Malaysia
*
Corresponding author
DOI:
https://dx.doi.org/10.47772/IJRISS.2025.910000006
Received: 29 September 2025; Accepted: 04 October 2025; Published: 01 November 2025
ABSTRACT
Mastering speaking skills in Arabic is a major challenge among students in Malaysia, as traditional teaching
approaches emphasize grammar and translation more than oral communication. This study evaluates the impact
of using ChatGPT AI Audio in enhancing s
pronunciation correction, and grammar feedback. A total of 60 public university students were selected as
respondents, using a mixed-methods design involving quantitative questionnaires, semi-structured interviews,
and written reflections. Descriptive analysis showed a high level of acceptance with an overall mean score of
4.12, while the speaking confidence item recorded the highest average (M=4.35). Thematic analysis identified
five main themes: increased confidence, phonetic awareness, technical constraints, cultural sensitivity, and
dependence on AI. The findings indicate that ChatGPT AI Audio can serve as an effective support tool to enhance
wareness, although it remains limited by digital infrastructure issues
and cultural understanding constraints. This study recommends the gradual integration of AI into Arabic
language pedagogy as a complement to traditional methods.
Keywords ChatGPT AI Audio, Arabic speaking skills, foreign language learning, mixed methods, and higher
education
INTRODUCTION
but also a language of knowledge, culture, and international diplomacy. Although many students study Arabic
at the primary, secondary, and even university levels, the greatest challenge often mentioned is the mastery of
speaking skills. Many students are able to read Arabic texts or understand grammar in theory, yet fail to convey
messages orally with fluency.
One contributing factor is the teaching approach that places greater emphasis on grammar (nahu), morphology
(sarf), and text translation rather than oral interaction (Al-Sobhi & Preece, 2018). This situation results in
students having fewer opportunities to actively practice speaking in Arabic. Furthermore, limitations in the
to imitate the correct intonation.
In the era of the Fourth Industrial Revolution (IR 4.0), the use of artificial intelligence (AI) has begun to attract
attention in the field of education. One of the latest innovations is ChatGPT with audio functions, which enables
real-time conversational
pronunciation, improves sentence structure, and suggests vocabulary in specific contexts. This technology has
the potential to reduce foreign language anxiety often experienced by students (Horwitz, 2016) and create
opportunities for more flexible self-directed learning.
This study was conducted to evaluate the effects of using ChatGPT AI audio in mastering Arabic speaking skills
in Malaysia. It focuses on two main aspects: quantitative analysis based on surveys of 60 respondents, and
qualitative analysis based on students
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 56
www.rsisinternational.org
LITERATURE REVIEW
A. Previous studies
Previous studies have shown that non-native learners often face difficulties in phonetics, intonation, and sentence
structure when speaking Arabic. Habash (2010) emphasizes that Arabic has a complex morphology system based
on root words, making it different from Malay and English. In addition, the wide variation of dialects adds to the
difficulty, as students are exposed to Modern Standard Arabic (MSA) in the classroom but hear Egyptian,
Levantine, or Gulf dialects through media (Alqudsi et al., 2014).
Foreign language anxiety has also been identified as a major obstacle. Horwitz (2016) notes that students often
feel afraid of making mistakes, which leads them to remain silent. This situation reduces their opportunities to
practice actively.
The integration of technology in language learning has been proven to enhance motivation and learning
effectiveness. Hashim and Yunus (2018) reported that the use of interactive applications in language learning in
provided them with opportunities for self-directed learning. In the
global context, AI technologies such as Google Translate, Microsoft Translator, and the more recent ChatGPT
offer new platforms that not only translate text but also support voice conversations. Figure 1 illustrates ChatGPT
AI Audio. Fig. 1 show ChatGPT AI Audio.
Fig. 1 ChatGPT AI Audio
One of its advantages is providing a safe interactive space where students can practice speaking without fear of
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 57
www.rsisinternational.org
making mistakes or feeling embarrassed when corrected. This helps to reduce foreign language anxiety, which
is usually a major barrier to speaking fluently (Horwitz, 2010). In addition, the audio capability of ChatGPT,
which provides immediate feedback on pronunciation and intonation, enables students to identify and correct
errors quicklysomething that is sometimes difficult to achieve in traditional classrooms (Habash, 2010). In
terms of flexibility, students are also free to use this technology anytime and anywhere, making learning more
accessible and less monotonous (Hashim & Yunus, 2018). Moreover, ChatGPT can simulate various real-life
situations such as making acquaintances, conversing in class, shopping, or discussing religious topics, thereby
making practice more meaningful and relevant to daily life (Al-Sobhi & Preece, 2018).
Gao et al. (2023), through the SeamlessM4T study, emphasized that AI-based multimodal models are becoming
increasingly effective in automatic translation and speech recognition. This opens up opportunities for
applications such as ChatGPT to be used more widely in language learning, including Arabic. Nevertheless,
specific research on the effectiveness of ChatGPT AI audio in mastering Arabic speaking skills in Malaysia is
still limited. However, the use of AI must also comply with Shariah principles, meaning that it should not involve
complete dependence nor replace the role of teachers (Rapa et al., 2024; Surtahman et al., 2025).
Theory
Communicative Language Teaching (CLT) is a language learning approach that emphasizes meaningful
communication as the primary goal. It stresses that language mastery is not measured solely through grammar
bility to use the language in real-life situations to convey
and understand messages (Richards & Rodgers, 2014).
In the context of this study, ChatGPT AI Audio can be regarded as a medium for implementing CLT because it
giving students the opportunity to practice speaking without fear of making mistakes, while also providing real-
time feedback on pronunciation and grammar. This interaction resembles authentic conversations, which are the
core of CLT.
In addition, CLT emphasizes a variety of communicative contexts such as social conversations, marketplace
transactions, classroom learning, and religious discussions (Littlewood, 2004). ChatGPT AI Audio enables the
simulation of these various contexts, thereby providing a learning experience that is more enjoyable, meaningful,
and authentic.
Therefore, the implementation of ChatGPT audio aligns with the principles of CLT, especially by providing a
space for communication practice that prioritizes fluency over accuracy, with corrections occurring naturally
through interactive feedback.
METHODOLOGY
This study employed a mixed-methods approach that combined both quantitative and qualitative analyses
(Creswell & Plano Clark, 2018; Creswell, 2014). A total of 60 students from a public university in Malaysia who
were enrolled in a communicative Arabic course were selected as respondents. The sample was chosen through
purposive random sampling, taking into account students with basic to intermediate levels of proficiency (Cohen
et al., 2018).
background, experience using ChatGPT audio, and perceptions of its effects on improving speaking skills. A
five-point Likert scale (1 = strongly disagree to 5 = strongly agree) was employed to measure the level of
satisfaction, as recommended in quantitative educational research (Creswell, 2014; Cohen et al., 2018).
For the qualitative data, semi-structured interviews and written reflections were employed. Students were asked
to share their experiences using ChatGPT audio in speaking practice, including challenges, benefits, and
suggestions for improvement. All qualitative data were analyzed using thematic analysis (Braun & Clarke, 2006;
Flick, 2014) to identify recurring patterns.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 58
www.rsisinternational.org
Data collection was conducted over four weeks. In the first week, students were briefed on the use of ChatGPT
audio and instructed to use it for at least 15 minutes daily. During the second and third weeks, students were
asked to use the audio conversation function in three stages: basic sentences, themed conversations, and
advanced discussions. In the fourth week, questionnaires were distributed and interviews were conducted.
Quantitative data were analyzed using descriptive statistics (Cohen et al., 2018), while qualitative data were
analyzed manually through open coding, theme categorization, and final theme determination (Braun & Clarke,
2006; Flick, 2014).
FINDINGS
A. Quantitative Analysis
The survey findings from TABLE I indicated a positive level of acceptance toward the use of ChatGPT AI audio.
The overall mean score was high (M = 4.12), suggesting that students welcomed the use of ChatGPT audio in
speaking practice. The highest-rated ite
However, 60% of respondents reported internet connectivity issues as the main limitation.
TABLE I QUANTITATIVE ANALYSIS
Item Statement
Mean (M)
Standard Deviation
% Agree/ High
AI audio increases my confidence to speak in
Arabic
4.35
0.58
86.7% (52 students)
AI audio helps me correct pronunciation and
phonetics more quickly
4.28
0.62
81.7% (49 students)
Conversation practice with AI is more
engaging compared to traditional methods
4.12
0.71
76.7% (46 students)
I can use AI audio independently without
teacher guidance
3.95
0.69
68.3% (41 students)
There are technical constraints
(internet/access) when using ChatGPT AI
audio
3.88
0.73
60.0% (36 students)
Qualitative Analysis
This study involved a total of 60 university students enrolled in a communicative Arabic course. The survey
results from TABLE II indicated a positive acceptance of using ChatGPT AI audio to enhance speaking skills.
Overall, the mean score was high at 4.12, suggesting that the majority of students agreed on the effectiveness of
this application in oral proficiency.
The item with the highest score was “AI audio increases my confidence to speak in Arabic”, with a mean of 4.35
and a standard deviation of 0.58. A total of 86.7% or 52 students expressed agreement. These findings
demonstrate that the use of ChatGPT audio has a positive psychological effect on students, particularly in
reducing the fear of making mistakes when speaking. Increased confidence is an important factor in foreign
y is a major barrier to
speaking confidence.
In terms of phonetics and pronunciation, the item “AI audio helps me correct pronunciation and phonetics more
quickly” obtained a mean score of 4.28, with 81.7% agreement (49 students). This indicates that audio interaction
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 59
www.rsisinternational.org
with AI enables students to hear the correct pronunciation and practice repeating it without feeling embarrassed.
-native Arabic learners lies
in the ability to accurately imitate phonetics.
Next, items related to motivation and enjoyment also showed positive effects. The statement “Conversation
practice with AI is more engaging compared to traditional methods” recorded a mean of 4.12, with 76.7%
agreement (46 students). This highlights that ChatGPT audio introduces interactive and enjoyable elements, in
contrast to traditional face-to-face learning, which is more formal and less flexible.
However, there were also findings that highlighted challenges. The item “I can use AI audio independently
without teacher guidance” obtained a mean of 3.95, with 68.3% (41 students) agreeing. While the majority of
students acknowledged their ability to learn independently, this data suggests that some still feel more confident
with teacher support. Meanwhile, technical issues were identified as the main limitation. The item “There are
technical constraints (internet/access) when using ChatGPT AI audio” recorded the lowest mean of 3.88, with
60% (36 students) acknowledging this. This indicates that digital infrastructure factors play a decisive role in
determining the effectiveness of this technology.
Overall, the quantitative results show that ChatGPT AI audio was well received by the students. It not only
enhanced their confidence but also proved effective in supporting phonetics, motivation, and opportunities for
self-directed learning. However, technical aspects and pedagogical support remain complementary factors that
need to be emphasized.
TABLE II Qualitative Analysis
Theme/Subtheme
F
%
Sample Respondent Quotes (Random R1R60)
Increased
Confidence
48
80.0%
Phonetic &
Pronunciation
Awareness
44
73.3%
, but with AI I dared to
and
, which
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 60
www.rsisinternational.org
, which is
and
is different from after
Technical
Constraints
36
60.0%
-
Literal Translation
/ Cultural Context
22
36.7%
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 61
www.rsisinternational.org
Dependence on AI
18
30.0%
-
1) Increased Confidence: This theme emerged as the most dominant, with 48 respondents (80%) reporting that
am not embarrassed to speak with AI even
a practice space free from the fear of embarrassment or ridicule. These results are consistent with the quantitative
= 4.35, SD = 0.58), with 86.7% of students agreeing. Clearly, affective aspects such as courage and self-
confidence constitute the most significant impact of this technology.
2) Phonetic and Pronunciation Awareness: A total of 44 respondents (73.3%) emphasized that AI helped them
of
and
is different from
4.28, SD = 0.62), with 81.7% of students agreeing. Indirectly, AI audio not only provides exposure to authentic
pronunciation but also builds stronger phonological awareness among Malay learners, who often struggle with
certain Arabic letters.
3) Technical Constraints: Despite the significant positive impact, 36 respondents (60%) reported technical issues
as the main challenge. Respondents mentioned problems such as weak internet connections, application freezes,
nternet is slow, the practice gets interrupted and I lose
cess) when using ChatGPT
although AI is effective as a learning tool, digital infrastructure remains a key barrier to ensuring a smooth
learning experience.
4)
phrases literally, whereas the real meaning
Although this issue was not highlighted in the quantitative survey, the qualitative findings add value by revealing
a deeper aspectthat AI still requires contextual adaptation to avoid misunderstandings, especially in the
translation of religious texts.
5) Dependence on AI: The final theme revealed that 18 respondents (30%) expressed concerns about the risk of
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 62
www.rsisinternational.org
overreliance and dependency on AI. Sample quotes include: “If I get too used to AI, I might feel less confident
speaking with real people” (R38) and “AI is suitable for initial practice, but not for every situation” (R40). This
highlights that while AI helps improve oral proficiency, it cannot replace real human interaction, which involves
body language, emotions, and social context. Therefore, AI should be used as a complement rather than a
substitute for traditional methods and authentic interaction.
This is the content and modules used for the conversation between students and AI audio on TABLE III.
TABLE III content and modules
Dialog
Arabic Text
Transliteration
English Translation
Notes
1: Greeting &
Introduction
Student:
AI:
Student:
AI:
.
-
yawm?
AI: Wa
anta?
anta?
Student: Peace be upon
you, how are you today?
AI: And peace be upon
you, I am fine, thank
you. And you?
Student: I am also fine.
My name is Ahmad, and
you?
AI: My name is Ali. I am
pleased to meet you,
Ahmad.
Basic practice
for greetings
and self-
introduction.
Suitable for
beginners.
2: In the
Classroom
Student:
AI:
.
Student:
.
AI:
al-
al- - -
Student: Shukran, al-
afhamu
Student: Teacher, can
you explain this lesson
again?
AI: Yes, of course. This
lesson is about past tense
verbs.
Student: Thank you,
now I understand better.
AI: Excellent! Do you
want additional
examples?
Academic
setting
dialogue.
Encourages
students to ask
questions and
understand
responses.
3: Shopping at
the Market
Student:
AI:
.
Student:
AI:
.
al-
ringgit.
Student: Hal yumkinuka
-
ringgit.
Student: How much is
this book?
AI: Its price is fifty
ringgit.
Student: Can you lower
the price a little?
AI: Okay, forty ringgit.
Practical
communicatio
n practice in a
market setting.
Trains
numbers and
bargaining
skills.
4: Basic
Religious
Student:
-maghrib?
Student: When do we
pray Maghrib?
Religious
context
dialogue.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 63
www.rsisinternational.org
Conversation
AI:
.
Student:
AI: .
ash-
AI: We pray right after
sunset.
Student: And how many
Links
language
learning to
daily worship
practices.
Dialogue 1, Greeting and Introduction: This dialogue focuses on the use of basic phrases in social encounters,
such as greeting, asking about well-being, introducing oneself, and giving polite responses. The student greets,
asks about well-being, and introduces their name, while the AI provides appropriate replies. The analysis shows
that this dialogue is highly relevant for beginner students as it emphasizes routine expressions used in daily
communication. The use of simple language structures, basic vocabulary, and greetings as an entry point to
it provides space for students to practice the intonation of greetings and the construction of short sentences,
which are the foundation of speaking skills.
Dialogue 2, In the Classroom: The second dialogue illustrates a formal situation in the classroom, where the
response contains a brief explanation of the grammar topic, namely the past tense, before the student expresses
their understanding. This exercise trains students to be confident in asking questions and expressing
understanding, two essential components of active learning. From a pedagogical perspective, it not only
emphasizes two-way communication but also sharpens the use of academic vocabulary such as “
” (lesson)
and “
” (past tense verbs). This situation familiarizes students with the instructional language they
will encounter in formal learning contexts.
Dialogue 3, Shopping at the Market: The third dialogue involves a practical situation, that is bargaining at the
market. The student asks for the price of a book, attempts to bargain, and accepts the new price offered by the
AI. From a linguistic perspective, the dialogue emphasizes the use of numbers, vocabulary related to transactions,
and the structure of price inquiries. This context is important because it connects language learning with real-life
situations. With such practice, students can master functional language for daily needs. In addition, this dialogue
encourages students to practice polite request intonation, which is highly relevant in Arab culture where
bargaining is a common social practice.
Dialogue 4, Basic Religious Conversation: The final dialogue touches on a religious theme, namely the time of
the Maghrib prayer and the number of rakaat. The AI provides a brief, accurate, and context-appropriate
response. This dialogue demonstrates that the integration of Arabic is not limited to social and academic
communication but also encompasses spiritual practices that are closely connected to the lives of Muslim
students in Malaysia. From an analytical perspective, this exercise is highly significant as it integrates language
mastery with religious appreciation. Students not only learn to pronounce words such as “
” (Maghrib) and
“
” (rakaat), but also understand the contextual use of the language in daily life. This aligns with the principle
of meaningful learning, where language is acquired within contexts that are relevant and valuable to students.
CONCLUSIONS
Overall, the findings of this study demonstrate that the use of ChatGPT AI audio has a positive impact on the
mastery of Arabic speaking skills among university students in Malaysia. These findings can be analyzed through
several key dimensions: self-confidence, phonetic mastery, learning motivation, technical constraints, as well as
cultural sensitivity and dependence on technology. This discussion is also supported by the findings of previous
studies outlined in the literature review.
Firstly, self-confidence emerged as the most dominant effect. A total of 80% of respondents reported that they
became more willing to speak after practicing with AI. This can be linked to the findings of Horwitz (2016), who
emphasized that foreign language anxiety is often the main barrier for students in foreign language
communication. ChatGPT audio, with its non-judgmental nature, provides a safe space to practice without social
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 64
www.rsisinternational.org
pressure. This phenomenon resembles the findings of Hashim and Yunus (2018), who found that the use of
functions not only as a technological tool bu
confidence.
Secondly, improvement in phonetic and pronunciation awareness also received strong support from the data. A
total of 73.3% of respondents stated that they became more sensitive to sound distinctions, for example
differentiating the pronunciation of
emphasizes the complexity of the Arabic phonological system based on root words, often posing challenges to
non-native speakers. The interactive audio practice offered by ChatGPT helps address this issue as it provides
instant input and can be repeated without limits. These findings are also consistent with Gao et al. (2023), who
demonstrated the effectiveness of AI multimodal models in speech recognition and translation.
Thirdly, in terms of learning motivation and enjoyment, the majority of respondents stated that conversation
practice with AI was more enjoyable compared to traditional methods. This is consistent with the findings of
Hashim and Yunus (2018), which reported that the integration of technology can enhance student engagement
in language learning. ChatGPT audio not only provides a variety of conversational contexts such as greetings,
classroom learning, shopping, and religious discussions, but also creates authentic experiences that are closer to
as emphasized by constructivist learning theory.
However, technical constraints are an issue that cannot be overlooked. A total of 60% of respondents reported
problems with internet access and application instability. This reflects the challenges of digital education
infrastructure in Malaysia, where the effectiveness of technology use is highly dependent on the quality of
infrastructure. Without stable internet connectivity, the advantages of AI audio cannot be fully utilized. These
findings indicate that education policymakers must give serious attention to improving digital access, particularly
in the context of language learning, which requires continuous interaction.
In addition, the theme of literal translation and cultural context also emerged from the qualitative analysis. Some
students felt that AI is still unable to fully grasp the cultural nuances of Arabic and the sensitivity of religious
contexts, particularly
al. (2014), who argued that dialectal variation and cultural context make Arabic more complex to understand
literally. This indicates that although AI works well as a technical training tool, it still requires human guidance
in addressing more subtle contextual aspects.
Finally, the issue of dependence on AI also emerged. Some respondents were concerned that excessive use of
AI could reduce their confidence in speaking with real speakers. This reminds us that technology should be
viewed as a support tool, not a complete replacement for teachers and human interaction. As noted by Al-Sobhi
and Preece (2018), the effectiveness of language learning depends on a balance between theory, practice, and
authentic interaction. Therefore, ChatGPT audio should be regarded as a complement to traditional teaching, not
as an absolute substitute (Rapa et al., 2024; Surtahman et al., 2025).
emphasizing the potential of technology in strengthening language learning. However, it also highlights that the
effectiveness of ChatGPT AI audio in mastering Arabic speaking skills cannot be separated from broader
psychological, technical, cultural, and pedagogical factors.
Limitation
Although this study provides positive insights into the use of ChatGPT AI audio in improving Arabic speaking
skills, there are several limitations that should be noted. The first limitation is that the study involved only 60
students from a public university in Malaysia. This relatively small sample size may not reflect the wider
population of Arabic language learners across other educational institutions, such as secondary schools, private
colleges, or religious learning centers. Therefore, the generalization of the findings should be made with caution.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 65
www.rsisinternational.org
Another limitation concerns the intervention period, which was restricted to only four weeks. Although the
-term effects of
using AI audio on communicative fluency and speaking proficiency cannot be determined. Further research with
a longer intervention period is needed to evaluate the sustained effectiveness more accurately.
A further limitation relates to technical factors. A large number of respondents reported unstable internet access,
which could disrupt the smooth use of the application. This indicates that the effectiveness of AI audio is highly
dependent on reliable digital infrastructure. In the context of rural areas or communities with limited internet
access, the effectiveness of using this technology may be lower compared to environments with high-quality
broadband connectivity.
the sample consisted of students with basic to intermediate levels, the analysis did not compare the effects of AI
use across groups with different proficiency levels. More focused future research could provide more specific
insights into how AI impacts learners differently according to their proficiency level.
Finally, the aspect of cultural and religious sensitivity was not examined in depth in the quantitative survey but
only emerged in the qualitative findings. Although this issue was not dominant, it remains important because
Arabic is closely tied to religious and cultural contexts. AI audio still has limitations in understanding these
nuances, and this requires further attention to ensure that the technology is not misinterpreted or used in
inappropriate contexts.
CONCLUSION
This study showed that the use of ChatGPT AI audio had a significant effect on the mastery of Arabic speaking skills
in Malaysia. More than 80% of respondents reported improvements in confidence and phonetic proficiency, while
technical issues, literal translation, and dependence on AI emerged as challenges. Therefore, several recommendations
are proposed. Firstly, educational institutions should provide clear guidelines for the use of AI in Arabic language
teaching. Secondly, the government and relevant authorities need to strengthen digital infrastructure to ensure stable
internet access. Thirdly, teachers should be trained to guide students in using AI in a balanced way, especially
concerning cultural and religious issues. Finally, further research should be conducted to evaluate the long-term effects
of using AI audio on Arabic language proficiency in Malaysia. In conclusion, ChatGPT AI audio opens new
opportunities in Arabic language pedagogy. It has the potential to serve as an effective support tool to
confidence, fluency, and phonetic awareness. However, its effectiveness depends on how it is integrated with
traditional methods, as well as on comprehensive infrastructural and pedagogical support.
ACKNOWLEDGEMENT
Thank you to University Technology MARA for the support of this research.
REFERENCES
1. Alqudsi, A., Omar, N., & Shaker, K. (2014). Arabic Machine Translation: A Survey. Artificial
Intelligence Review, 42(4), 549572.
https://doi.org/10.1007/s10462-012-9351-1
2. Al-Sobhi, B. M. S., & Preece, A. S. (2018). Teaching English Speaking Skills To The Arab Students In
The Saudi School In Kuala Lumpur: Problems and Solutions. International Journal of Education and
Literacy Studies, 6(1), 111.
3. Barrault, L., Elbayad, M., Mani, A., Antonios, S., Pino, J., & Meta AI Research. (2023). Seamlessm4t:
Massively Multilingual and Multimodal Machine Translation. arXiv preprint arXiv:2308.11596.
https://arxiv.org/abs/2308.11596
4. Braun, V., & Clarke, V. (2006). Using Thematic Analysis In Psychology. Qualitative Research In
Psychology, 3(2), 77101.
https://doi.org/10.1191/1478088706qp063oa
5. Cohen, L., Manion, L., & Morrison, K. (2018). Research Methods In Education (8th ed.). Routledge.
https://doi.org/10.4324/9781315456539
6. Creswell, J. W. (2014). Research Design: Qualitative, Quantitative, And Mixed Methods Approaches
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 66
www.rsisinternational.org
(4th ed.). SAGE Publications.
7. Creswell, J. W., & Plano Clark, V. L. (2018). Designing And Conducting Mixed Methods Research
(3rd ed.). SAGE Publications.
8. Flick, U. (2014). An Introduction To Qualitative Research (5th ed.). SAGE Publications.
9. Habash, N. Y. (2010). Introduction To Arabic Natural Language Processing. Morgan & Claypool.
https://doi.org/10.2200/S00277ED1V01Y201008HLT010
10. Hashim, H. U., & Yunus, M. M. (2018). English As A Second Language (ESL) Learning: Setting The
Right Environment For Second Language Acquisition. Tadris: Journal Keguruan Dan Ilmu Tarbiyah,
3(2), 207215.
https://doi.org/10.24042/tadris.v3i2.2941
11. Horwitz, E. K. (2010). Foreign And Second Language Anxiety. Language Teaching, 43(2), 154167.
https://doi.org/10.1017/S026144480999036X
12. Horwitz, E. K., Horwitz, M. B., & Cope, J. (1986). Foreign Language Classroom Anxiety. The Modern
Language Journal, 70(2), 125132.
https://doi.org/10.1111/j.1540-4781.1986.tb05256.x
13. Littlewood, W. (2004). The Task-Based Approach: Some Questions And Suggestions. ELT Journal,
58(4), 319326.
https://doi.org/10.1093/elt/58.4.319
14. Rapa, A. A., Saja, I., & Azmi, A. (2024). The Use Of Artificial Intelligence (AI) Translation Tools:
Implications For Third Language Proficiency. International Journal Of Research And Innovation in
Social Science, 8(9), 1952-1960.
15. Richards, J. C., & Rodgers, T. S. (2014). Approaches And Methods In Language Teaching (3rd ed.).
Cambridge University Press.
16. Impact
Kecerdasan Buatan (AI) dalam Sibergogi Pendidikan Sepanjang Hayat daripada Perspektif Islam dan
Barat [The Impact of Artificial Intelligence (AI) in Cybergogy of Lifelong Learning from Islamic and
Western Perspectives]. Global Journal Al-Thaqafah. 15. 189-208.
https://doi.org//10.7187/GJAT072025-12
